Received September 4, 2020, accepted September 8, 2020, date of publication September 11, 2020,

date of current version September 24, 2020.

Digital Object Identifier 10.1109/ACCESS.2020.3023473

Protect Your Sky: A Survey of Counter Unmanned

Aerial Vehicle Systems
HONGGU KANG 1 , (Student Member, IEEE), JINGON JOUNG 2 , (Senior Member, IEEE),
JINYOUNG KIM3 , (Member, IEEE), JOONHYUK KANG 1 , (Member, IEEE),
AND YONG SOO CHO 2 , (Senior Member, IEEE)
1 Department of Electrical Engineering, KAIST, Daejeon 34141, South Korea
2 School of Electrical and Electronics Engineering, Chung-Ang University, Seoul 06974, South Korea
3 Korea University Business School, Korea University, Seoul 02841, South Korea
Corresponding author: Jingon Joung (jgjoung@cau.ac.kr)
This work was supported in part by the National Research Foundation of Korea (NRF) Grant funded by the Korea Government (MSIT)
under Grant 2018R1A4A1023826, and in part by the Ministry of Science and ICT (MSIT), South Korea, through the Information
Technology Research Center (ITRC) Support Program supervised by the Institute of Information and Communications Technology
Planning and Evaluation (IITP) under Grant IITP-2020-0-01787.

ABSTRACT Recognizing the various and broad range of applications of unmanned aerial vehicles (UAVs)
and unmanned aircraft systems (UAS) for personal, public and military applications, recent un-intentional
malfunctions of uncontrollable UAVs or intentional attacks on them divert our attention and motivate us to
devise a protection system, referred to as a counter UAV system (CUS). The CUS, also known as a counter-
drone system, protects personal, commercial, public, and military facilities and areas from uncontrollable
and belligerent UAVs by neutralizing or destroying them. This paper provides a comprehensive survey
of the CUS to describe the key technologies of the CUS and provide sufficient information with wich to
comprehend this system. The first part starts with an introduction of general UAVs and the concept of
the CUS. In the second part, we provide an extensive survey of the CUS through a top-down approach:
i) the platform of CUS including ground and sky platforms and related networks; ii) the architecture of
the CUS consisting of sensing systems, command-and-control (C2) systems, and mitigation systems; and
iii) the devices and functions with the sensors for detection-and-identification and localization-and-tracking
actions and mitigators for neutralization. The last part is devoted to a survey of the CUS market with relevant
challenges and future visions. From the CUS market survey, potential readers can identify the major players
in a CUS industry and obtain information with which to develop the CUS industry. A broad understanding
gained from the survey overall will assist with the design of a holistic CUS and inspire cross-domain research
across physical layer designs in wireless communications, CUS network designs, control theory, mechanics,
and computer science, to enhance counter UAV techniques further.

INDEX TERMS Unmanned aerial vehicle (UAV), unmanned aircraft system (UAS), counter UAV system
(CUS), counter drone systems, public safety, defense.

I. INTRODUCTION cases of UAVs, also widely known as drones. In this paper,

Given the various practical and potential applications and pur- UAV is used as a general term for unmanned aircraft, includ-
poses behind the use of unmanned aerial vehicles (UAVs) and ing remotely piloted aircraft controlled by an operator on the
unmanned aircraft systems (UAS) from non-public hobbies ground and drones that can fly autonomously [1]. Though
to military purposes, UAVs and UASs have been rigorously the word ‘drone’ can be used to describe a wide variety of
studied and developed over the last 30 years. Currently, in real vehicles, including even seafaring submarines and land-based
life, we can readily observe various public and non-public use autonomously vehicles, UAVs (or UASs) and drones are used
interchangeably throughout the paper. Moreover, the UAS,
The associate editor coordinating the review of this manuscript and which consists of a UAV and the controllers, is also used
approving it for publication was Chunsheng Zhu . interchangeably with UAV.

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
VOLUME 8, 2020 168671
 H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

A. MOTIVATIONS The White House has not remained exempt from threats
1) RAPID GROWTH OF UAVs AND THEIR APPLICATIONS of rogue drones either; a DJI quadcopter for recreational
Applications of UAVs range from recreation to commercial purposes accidentally crash-landed on the south lawn of the
and military applications, including enjoyment, hobbies, and White House in 2015 [33]. The benign nature in these cases,
games with drones, the filming of movies for recreation [2]– however, was not replicated in subsequent incidents. About
[4], and the operation of UAVs for military purposes [5]–[11]. a year and a half later, a Japanese protester against the use
As reported by the federal aviation administration (FAA) in of nuclear power managed to land a drone, marked with an
the United States (US), there are 1, 692, 700 registered drones odious radioactive sign, on the roof of the Japanese prime
(approximately, 29% for commercial and 71% for recreation) minister’s office [34]. The drone was carrying a container
in the US as of September 2020 [12]. The commercial UAV filled with radioactive sand from Fukushima. Multiple major
industry, whose dynamics has been dubbed a modern-day news outlets have started to voice serious concerns over hos-
gold rush by multiple industry players [13], has grown rapidly tile drones (e.g., [35], [36]). Recently, hostility by malignant
in tandem with expanding market needs, such as disaster drones became apparent to the general public when Nicolas
management, emergency services, agricultural applications, Maduro, the President of Venezuela, was attacked by two
cargo inspection, or recreational purposes, to name a few. commercial drones, each of which contained one kilogram
Clearly, UAVs are considered as an essential enabler to of C-4 explosive, in Caracas, Venezuela, in August of 2018
enlarge commercial markets and are used in various indus- [37]. This series of drone attacks on a head of state captured
tries, such as in i) the agricultural industry for seeding, cross- only a fraction of the negative externalities of the booming
pollination, and crop-dusting [14], [15]; ii) the distribution UAV industry. Rogue drones hovering over airports or private
industry for the delivery and/or collection of packages [16]– compounds pose diverse ranges of threats from security to
[19]; iii) the construction industry for building and measuring privacy. Stealth drones deliver contraband by dropping pack-
[20], [21]; and, iv) the information technology (IT) industry ages onto prison grounds. Concerns over potential threats by
for enlarging service coverage areas and establishing emer- UAVs have materialized quickly.
gency networks [22]–[27]. Furthermore, UAVs are used to Furthermore, according to a survey of online news articles,
provide effective public services, such as environmental (e.g., there were more than 200 incidents (100 in North Amer-
traffic and air pollution) monitoring [28]–[30] and firefight- ica, 77 in Europe, 38 in Asia/Pacific, 17 in the Middle
ing and rescue operations [31]. East, and 6 in Latin America) in 2019 [38]. As also shown
in Fig. 1, the number of incidents that are caused by UAVs,
as reported to the FAA in the US [12], generally increases
every year. Compared to the total number of incidents in
2015, i.e., 1, 213, this number increased by 76% to 2, 142 in
2019.

FIGURE 1. Number of incidents caused by UAVs in the United States from

January of 2015 to December of 2019, as reported to the FAA in the US
[12].

2) RAPID GROWTH OF ACCIDENTS AND CRIMES INVOLVED

FIGURE 2. Industries affected by UAV incidents across the globe,
IN UAVs as reported in online news articles between December of 2018 to
With the various and vigorous promising applications of March of 2020 and collected in earlier work [39].

UAVs, now is a suitable time to consider UAVs from a differ-

ent angle considering the possibility that they may threaten On the other hand, because UAVs have multidirectional
our safety. At a 2013 campaign rally in Dresden, Germany, purposes, their negative effects are also extensive. For exam-
a quadcopter drone hovered within a few feet of Angela ple, UAVs disturb current aviation operations, invade per-
Merkel, the Chancellor of Germany, and Thomas Maiziere, sonal privacy, and threaten public and national safety. Based
the German Defense Minister, eventually crashing in front on data, collected from online news articles between Decem-
of Merkel [32]. This harmless stunt was found to have been ber of 2018 and March of 2020 [39], the industries affected
orchestrated by the Pirate Party in the form of a protest against by the UAV-related incidents were determined These are
drone observation and government surveillance in Germany. categorized in Fig. 2. As verified in the analysis, UAVs

168672 VOLUME 8, 2020

H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

affect various industries; in particular, the majority of inci- TABLE 1. Relevant Surveys and Studies on CUS.
dents occur at airports, at a rate of approximately 35%.
An accident at an airport can cause serious disasters and even
fatalities, posing therefore a threat to both human life and
property.

3) LACK OF STUDIES AND SURVEYS ON COUNTER UAV

SYSTEMS
To mitigate such alarming effects caused by UAVs, govern-
ments regulate UAV operations via civil aeronautics laws
irrespective of the operators and operation [40]–[44]. For
the operators, a legal license, insurance, and registration are
required. For example, 171, 744 licenses have been issued
as of March of 2020 in the US [12]. With regard to oper-
ational regulations, an authorized private/public office con-
trols and restricts UAV operations by setting limits on the
maximum operation speed and height, locations, behaviors, B. ORGANIZATION AND CONTRIBUTIONS OF THIS
and communication frequency bands. However, because the SURVEY
current regulation passively controls UAV operations, it does
The acronyms frequently used in the paper and the taxonomy
not guarantee privacy and safety from uncontrollable UAVs,
of our survey on the CUS are shown in Tables 2 and 3,
e.g., those with unintentional malfunctions owing to a con-
respectively. Table 3 contains five columns for the survey
nection loss by the operator and the UAVs of the illegal
topics, subtopics, category, examples/descriptions with pros
intruders who attempt to attack public and military facili-
and cons, and references. The main survey consists of three
ties. In keeping with the rapid development of UAV tech-
parts from Section II to Section VI with the following five
nologies, the resultant threat from uncontrollable UAVs is
topics: i) UAV applications and regulations; ii) platforms and
inevitable. Therefore, to secure personal privacy, commercial,
networks of the CUS; iii) the CUS architecture; iv) devices
public, and military facilities and areas from uncontrollable
and functions of the CUS; and v) markets, challenges, and
and belligerent UAVs, i.e., malicious UAVs (mUAVs),1 a
the future vision of the CUS.
protection system, referred to here as a counter UAV system
(CUS), also known as (a.k.a.) a counter-drone system [38], • The first part, i.e., Section II, is an introductory part
is desired. that briefly introduces the various UAV applications
Compared to the regular aircrafts, the UAVs, in gen- and regulations for operators and operations. Moreover,
eral, have unique characteristics. For example, UAVs are the necessity of the CUS is justified by introducing the
unmanned, inexpensive/affordable, fly at low altitudes with concept of the CUS.
slow speed, and have limited payload. Therefore, the UAVs • In the second part, the CUS is rigorously surveyed
can reasonably (re)modeled to mUAVs, and the mitigators throughout Sections III, IV, and V through a top-down
against mUAVs are required to be studied separately from approach from the platform to the architecture of the
the existing studies for the defense of the regular airplanes. CUS followed by the devices and functions of the
The in-depth and large-scale surveys of CUS, however, are CUS. In Section III, the CUS platform is introduced
lack and the current surveys have been performed covering and categorized into three parts based on the opera-
only a part of CUS as summarized in Table 1 [45]–[52]. tion methodologies as follows: i) the ground platform,
On the other hand, in this survey, we provide a comprehen- i.e., the main platform accounting for approximately
sive survey for designing holistic CUS that includes plat- 90% of CUS platforms and operated on the ground in
form, architecture, devices, and their functions for CUS. The static, mobile, and handheld manners; ii) the sky plat-
CUS platforms will be categorized according to the mobility form, approximately 10% of all platforms and operated
and operating area. The CUS architecture including various at low or high altitudes; and iii) the CUS networks that
sensing, command and control (C2), and mitigation systems link multiple platforms. In Section IV, the CUS architec-
will be surveyed with the specific functions and devices. ture is surveyed with related topics, specifically, sensing
Furthermore, the challenges and vision of the related market systems, command-and-control (C2) systems, and mit-
will be provided. To the best of our knowledge, this study is igation systems. The sensing systems gather data from
the first work that comprehensively surveys on the CUS as the environments. The C2 systems perform computing
summarized in the following subsection. tasks, such as detection, identification, tracking, and
localization, and determine false alarms, establishing
whitelists/blacklists, and setting neutralizing methods
1 Throughout the paper, uncontrollable and belligerent UAVs, including according to the threat level. The mitigation systems
intrusion UAVs and hostile UAVs, are referred to as mUAVs. neutralize mUAVs. Section V introduces the sensors

VOLUME 8, 2020 168673

 H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

TABLE 2. Acronyms / Abbreviations Used in This Paper (Alphabetic order).

TABLE 3. Overview of the Survey.

and mitigators with their functions performed by the sensors, light detection and ranging (LiDAR), electro-
architecture topics of CUSs at the device level. Here, optical (EO)/infrared (IR) sensors, sound navigation
various sensors, such as radar, radio frequency (RF) ranging (sonar), and acoustic/ultrasonic sensors, and

168674 VOLUME 8, 2020

H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

mitigation devices (including lasers, projectiles, colli- acquisition, reconnaissance (ISTAR), combat, and commu-
sion UAVs, jammers, electromagnetic pulses (EMPs), nications [5]–[11], [53]–[56]. UAVs equipped with multi-
spoofing/hacking devices, and nets) are briefly surveyed ple sensors, e.g., EO/IR and acoustic sensors, can com-
and their limitations and requirements for the CUS are plete important reconnaissance and surveillance missions.
discussed. Exploiting multiple sensors on UAVs, useful information
• In the last section, the current trends and distinctive can be collected during surveillance, target acquisition, and
characteristics of the CUS market are identified to help reconnaissance, and can then be processed to make the better
readers understand the industrial and geographical dis- battle plans, with one example being ISTAR. Using advanced
tributions of the current CUS market, both the drivers communication technology, multiple drones can cooperate to
and inhibitors of the CUS market growth, and the major complete a military mission, such as video reconnaissance
players and their competitive yet cooperative dynamics [56]. Exploiting the relatively small form factor of mini UAVs
in the CUS market. Also, highly distinctive characteris- compared to a human-scale aircraft enables concealable
tics of the CUS market are identified as the asymmetric countermeasures such as radar and communication jammers
interdependence between the UAV and CUS industries, [10]. In addition, a mid-size unmanned combat aerial vehicle
the temporal precedence of the UAV industry, and the (UCAV) or a combat drone can carry aircraft ordnance,
complete dependence of the CUS market demise on such as missiles and/or bombs, and can be used for drone
both regulatory changes and the growth rate of the UAV strikes [53]. For an effective attack, the sufficient accuracy of
industry. detection and identification of the target location, i.e., target
The broad understanding gained from this survey will acquisition, are required. The small UAV can also be used
help design a holistic CUS to neutralize/destroy mUAVs to detect and eliminate land mines. Additionally, a UAV
and mitigate this threat by aspiring cross-domain research that operates as a base station (BS) or relay station can
across physical layer designs in wireless communications, enlarge the communication coverage area on a battlefield,
UAV network designs, control theory, mechanics, and com- where a BS is unavailable, such that emergent and short-time
puter science. communications become possible [5], [54]–[56].

II. UAV APPLICATIONS AND REGULATIONS

Recently, the commercial UAV market has grown gradually 2) CIVILIAN-NONCOMMERCIAL APPLICATIONS
and applications of UAVs have broadened from their typical The civilian-noncommercial applications of a UAV cover a
military purpose to various purposes as the cost of UAV sys- wide area, from public services to scientific research [28]–
tems decreases. With the explosive growth of UAVs, injuries [31], [57]–[60], [81]–[83].
and manual physical labor in the military and in industry • Monitoring: A UAV can fly and hover around hard-to-
have been reduced, and various leisure activities have newly access (H2A) or dangerous places, where monitoring is
appeared given the mobility and flexible operations of UAVs. necessary for safety, comfort, and scientific purposes.
On the other hand, the increased number of UAV applications For example, government facilities and public infras-
has also caused concern about potential accidents and crimes. tructure elements covering a wide area are challeng-
To prevent the misuse of UAV systems and illegal operations, ing for a person to monitor completely using a fixed
regulations pertaining to UAVs have been established in many camera or a simple patrol strategy. UAVs, however, can
countries. However, such regulations have a fundamental monitor a complete area at a glance from the sky or a
limit in that they cannot actively control accidents and ille- specific spot while flying at that spot, meaning that they
gal operations, and the potential threat of mUAVs remains. can cover a target area without blind and/or occluded
Therefore, an active defense system, i.e., the aforementioned spots. UAVs can patrol to monitor and detect instances
CUS, is desired. In this section, applications of UAVs and of spontaneous combustion at a relatively low cost. They
pertinent regulations are briefly introduced to clarify the can also be applied to the real-time monitoring of vehicle
motivation of this CUS survey, followed by the concept of the density levels to collect traffic information [29], [81].
CUS. A comprehensive survey of the CUS will be provided Note that conventional fixed surveillance cameras can
after this section. monitor only a part of the road. Though helicopters
A. UAV APPLICATIONS with cameras can obtain footage of roadways much
The main applications are categorized into military appli- more freely, the operation cost is extremely high. UAVs
cations, civilian-noncommercial (i.e., public) applications, can resolve such mobility and operation cost issues and
and civilian-commercial applications, including industry and collect useful information, such as detour routes, so that
personal applications, as summarized in Table 4. Various drivers can avoid traffic jams or accidents. In addition,
applications in each category are introduced below. UAV monitoring can also be used for scientific purposes,
e.g., for air pollution measuring [30] and for monitoring
1) MILITARY APPLICATIONS the status of active volcanos [28].
UAVs have been deployed in various military mis- • Relief activities: Other important public applications of
sions/operations, such as intelligence, surveillance, target UAVs are relief activities. Bulky pieces of equipment,

VOLUME 8, 2020 168675

 H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

TABLE 4. Applications and Functions of UAVs.

such as helicopters and fire trucks, cannot easily reach a • Agriculture: To increase crop yields, UAVs can assist in
building/house fire in urban areas due to various obsta- farming industries or can help farmers complete various
cles and traffic. Moreover, firefighters may not even be tasks, such as soil and field analyses, seeding, planting,
allowed to enter the building/house due to the possibility monitoring crop growth, cross-pollination, irrigation,
of a fatal collapse. Under such an emergent situation, health assessments, and crop-dusting [14], [15], [61],
a UAV can be serve as a lifesaver owing to its small [62]. Here, an essential technology enabling many of
size and mobility, which enables it to readily enter such the agricultural tasks of UAVs is the sensing capabil-
buildings, quickly investigate the situation, and report ity of UAVs. By detecting and tracking topographical
the circumstances inside. A UAV equipped with the and geographical variations using EO/IR sensors and
ability to spray water can also extinguish fires at critical LiDAR, UAVs can avoid collisions and create effi-
spots, such as at gas tanks and ignition points, imme- cient schedules of flight routes. Moreover, various sen-
diately without direct human control [31]. A UAV can sors, such as hyperspectral, multispectral, or thermal
also conduct rescue missions by probing H2A places, sensors, are required to monitor humidity levels and
reporting the locations of accidents, and comforting vic- temperatures.
tims after spotting them [59]. A fumigator UAV to fight • Construction: UAVs have already begun to be used in
pandemics and epidemics is another important relief the construction industry, reducing much human effort
activity of a UAV. For instance, fumigator drones were as well as errors associated with traditional constructing
deployed to prevent the spread of diseases, i.e., such as tasks [20], [21]. For example, UAVs can survey land
the coronavirus (COVID-19) in South Korea early 2020 from the perspective of drones, monitor the safety of
[60]. Fumigator drones spray disinfectants over a vast the laborers, protect construction sites from theft or van-
area in a short time, requiring the least amount of man- dalism, inspect numerous dangers and safety hazards
power. Moreover, disinfectants sprayed via a UAV can through three-dimensional (3D) mapping, and provide
easily fumigate blind spots that are normally difficult to video footage to facilitate communications and surveil-
reach by human hands. Note that using fumigator drones lance. In such cases, along with the sensing capability,
for the prevention of epidemics is controversial because which is the essential technology of agricultural UAVs,
the effectiveness of this strategy depends on the type of the communication capability should be emphasized as
virus and whether the contagion can spread aerially, yet an essential technology as well, enabling the advantages
fumigator drones will be further developed and widely of UAVs at construction sites. Very low-latency commu-
used owing to their potential benefits in this area. nication is essential for construction UAVs to prevent
accidents at construction sites. To this end, 5G/beyond
3) CIVILIAN-COMMERCIAL APPLICATIONS 5G (B5G) technology can be applied to these UAVs. The
Various industries from large companies to small start-up 3GPP Working Groups ensure that the 5G system will
companies exploit the benefits of UAV to increase their profit. meet the connectivity needs of UASs [84]. Considering
Although the commercial/industrial use of UAVs is relatively the UAVs as an invaluable tool in construction, UAVs
new compared to military uses, there are a wide variety of will take on even more integral and complex tasks asso-
civilian-commercial applications [82], [83]. ciated with large projects in the future.

168676 VOLUME 8, 2020

H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

• Delivery service: UAVs can be used to transport • IT services: As one of the most promising applications
lightweight medicines and vaccines, packages, food, and of commercial UAVs is to provide IT services, where
other small goods into or out of remote or otherwise the UAV operates as, for example, a BS, relay, and/or
inaccessible regions (i.e., an H2A region). For exam- data corrector from sensors, to enhance the quality of
ple, UAVs can transport medicines and vaccines into IT services. Especially in relation to wireless communi-
H2A regions [63], [64]. They can also retrieve med- cations, there have been many comprehensive surveys
ical samples from an H2A region. Many postal com- of how wireless communications can be enhanced by
panies from the US, Australia, Switzerland, Germany, UAVs (e.g., [25], [43], [67]–[71], [73]–[77], [80] for
Singapore, and Ukraine have tested the feasibility and communication applications aided by UAVs and the
profitability of courier services using UAVs [65]. Food references therein.). Examples include broadband com-
delivery UAVs, specifically rotary-wing types, have also munications [67], internet-of-things (IoT) applications
been demonstrated by many companies involved in the [70], [72], communication platforms depending on the
foodservice industry. Because UAVs are a power-limited altitude of UAVs [74], wireless channel models involved
system, to complete their delivery services given their in UAV communications [75], [76], cellular systems
limited battery power or fuel (e.g., Amazon ‘Prime Air’ supported by UAVs [43], [77], and data links [80].
carrying a package up to approximately 2 kg with a 13- To enhance the many wireless communication appli-
min flight time to the destination [85]), delivery path cations, rigorous and various, technical and theoretical
optimization and UAV status monitoring methods have studies have been conducted to find the optimal designs
been studied [16]–[19]. Before the expected widespread of the parameters involved in UAV communications,
usage of UAVs as courierr in the future, appropriate such as the trajectory and placement of UAVs [22]–[24],
regulations should be established to overcome safety [26], [27], [78], [79], resource usage (e.g., power and
and legal hurdles and prevent their potential illegal use, time) [27], [87]–[89], and proper topologies [90], [91].
as reported in Section I.
• Recreation: Diverse UAVs ranging from low-cost toys B. REGULATION PERTAINING TO UAV OPERATIONS
to expensive high-end products for civilian applications As introduced in the previous subsection, numerous appli-
such as filmmaking, photography, racing, and com- cations of UAVs have been introduced or will eventually
mercial advertisements, are easy to find in society at be introduced, with enormous benefits. Various incidents,
present. Depending on the application type, many key however, accompanied by the increase in UAV-aided services
technologies are involved. Controlling the 1,218 UAVs and technologies will also increase, as stated in Section I.
performing the light show at opening ceremony of the To prevent unwanted incidents caused by UAVs, regula-
Olympic Winter Games PyeongChang, South Korea, tions on commercial UAVs have been established in many
in 2018 required seamless control technology and com- countries [40]–[43], [80], [92]. The details of these regu-
munications technology to provide the massive number lations vary from country to country. For example, a pilot
of connections between the UAVs and a control center license is mandatory for operation in some countries, e.g.,
to keep them all airborne simultaneously [86]. Taking the US, China, and the United Kingdom (UK), though not
video and photos using UAVs requires stabilizer tech- all. In South Korea and Australia, a pilot license is required
nology to obtain a clear shot from the UAVs [2]. In addi- only if the weight of the drone exceeds a specified standard.
tion, customizing the software and hardware of UAVs, The aviation authorities of 132 countries all across the globe
as is done with what are termed do-it-yourself (DIY) have also created regulations [44]. Although regulations vary
UAVs, and flying and controlling UAVs during races widely among countries, their common purpose is to prevent
have become a type of e-sport recently. For example, unwanted incidents stemming from UAV operations, and they
in the global drone racing league MultiGP, which started can be categorized into regulations pertaining to operators
in 2015 [66], a pilot controls the UAV by observing and those affecting operations, as shown in Table 5.
footage from a camera mounted on the UAV with the sig-
nal sent to goggles or a monitor worn by the pilot, i.e., a 1) REGULATIONS ON OPERATORS
first-person view (FPV) or ‘video flying’. Here, efficient UAV operators in many countries are regulated by laws in
image processing and communication technologies are their countries. Specifically, a pilot license and insurance are
required for seamless and high-quality video streaming required under specific environments or in all cases in some
(typically a frequency of 2.4 GHz or 5.8 GHz). Like countries, such as Australia, where a pilot license is required
a traditional robot maze competition, a UAV race can if the weight of the UAV exceeds two kilograms. Likewise,
serve to evaluate and validate a learning algorithm to in the US, the pilot license is required (mandatory for com-
determine optimal paths in the sky [3], [4]. For personal mercial purposes) and a re-evaluation of pilot competency
recreation purposes, the pilots of UAVs should recognize should be conducted every two years. Moreover, pilot training
and follow the regulations and practice basic courtesy to is required for beyond–visual–line–of–sight (BVLoS) opera-
ensure public safety and privacy. tions in some countries.

VOLUME 8, 2020 168677

 H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

TABLE 5. Regulations on Commercial UAVs in 20 countries and Violation Consequences [40]–[44].

2) REGULATIONS ON OPERATION UAV flies too far away under a BVLoS environment,
UAV regulations specify certain operational constraints, such the strength of the communication signals becomes
as maximum speeds, maximum heights, minimum distances insufficient and the pilot may lose control. In these
regarding certain areas or objects, approved flight areas and cases, the UAV can intrude upon private property or any
behaviors, and set operating frequency bands. Most coun- restricted area and can result in casualties and/or prop-
tries regulate the maximum height and speed of UAVs. The erty damage. There is a high probability that such acci-
minimum distances to people, vehicles, or certain areas such dents occur when the pilot is unqualified. Unless the
as military bases is also specified. In some countries, only pilot has a license or the UAV is registered with appro-
a visual–line–of–sight (VLoS) between the UAV and the priate insurance, tracking a suspect is also difficult, and
operator is allowed during UAV operation; i.e., the UAV this causes a delay of the recovery process. Note that
operation under the BVLoS is not allowed, as unclear sight approximately 70% of the incidents shown in Fig. 2 were
may cause an incident with high probability while operating caused by such an intrusion.
UAVs. However, some countries allow BVLoS operation if a • Non-violent crimes: Violating regulations, a UAV could
collision-avoidance function is employed by the UAV. UAV be misapplied and used for non-violent crimes, such as
registration is required in some countries. During UAV com- privacy intrusions, data robberies, and illegal deliveries.
munications, a data link should be established within a pre- Specifically, an offender could attempt to gather pri-
determined frequency band according to certain regulations. vate or secret information from civilians, officers or ser-
Regulations also define basic ethical courtesies carrying vicepersons by taking photographs and eavesdropping
no legal binding force to protect privacy and safety, e.g., on them. Conveying illegal objects such as unauthorized
no flying over private property, no carrying of hazardous firearms, explosives, and drugs could also be conducted
materials, and no dropping of any item. using an unauthorized UAV. For example, as shown
in Fig. 2, there were several crimes accounting for more
3) REGULATION VIOLATIONS than 10% among incidents to smuggle contraband into
Though regulations of UAV systems have been established to prisons.
prevent incidents, they passively control the potential misuse • Violent crimes: violent crimes, i.e., attacks, directly
of UAVs and can be violated intentionally or unintentionally. threaten our safety with possibly fatal outcomes. Vio-
Thus, violating a regulation and the consequent effects should lent crimes are closely related to political and military
be clearly understood and examined to develop appropriate issues, such as terrorism, and are relatively rare com-
countermeasures so that the remaining threats to private pri- pared to accidental and non-violent crimes. However,
vacy and public safety can be reduced further. To this end, as UAVs become more easily accessible to the pub-
violations of regulations and the accompanying results are lic, there is growing apprehension that violent crimes
categorized into three different cases, with possible counter- involving them will increase.
measures and technologies. To prevent possible damage from accidents, non-violent
• Accidents: The regulations on maximum heights or crimes, and violent crimes with mUAVs, further clear and
speeds can be violated unintentionally owing to a lack concrete regulations are required. Hence, both regional and
of caution or unexpected disturbances such as wind. If a international regulations pertaining to UAVs continue to be

168678 VOLUME 8, 2020

H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

established. Furthermore, for safety and to protect our prop- C2 systems are partially distributed to each platform. The
erty from mUAV misapplications and to enjoy the enormous details of C2 systems are discussed in Section IV.
benefits from various UAV applications, further active coun-
termeasures that effectively detect and mitigate mUAVs are A. GROUND PLATFORM
necessary. Henceforth, a comprehensive survey of defense
Ground platforms are classified as the static ground, mobile
systems is provided.
ground, and human-packable platforms according to the oper-
ation method. Static ground platforms are typically heavy
III. PLATFORMS AND NETWORKS OF CUS’s and thus are deployed and operated at a fixed location.
As stated in the previous section, a defense system is required On the other hand, mobile ground platforms are typically
for the active protection our safety, property, and prosper- vehicle-mounted that can be operated on the move or at a
ous future life. Defense systems to prevent unwanted inci- fixed location. Human-packable (handheld/wearable) plat-
dents, crime, and attacks from the misapplication of UAVs, forms are compact and portable so as to be carried and oper-
i.e., mUAVs, are referred to as CUSs. A CUS detects, rec- ated by a human. The details of each platform are surveyed
ognizes, tracks, and mitigates mUAVs. Moreover, a CUS can below.
localize the pilot of an mUAV. In this section, the details of It is worth noting that, following characteristics of the
CUSs will be surveyed based on their platforms and networks. CUS platforms, a game-theoretic problem can be formulated
We categorize the platforms of CUSs into the two classes between mUAVs and CUS. The CUS tries to restrict and deter
of ground and sky platforms, as illustrated in Fig. 3. Ground mUAVs, whereas the mUAVs attempt to complete their mis-
and sky platforms consist of CUSs that operate on the ground sions (e.g., reaching destination to perform harmful behav-
and in the sky, respectively. Ground platforms can further ior). The mUAVs may try to find a path that is not the shortest,
be classified into static ground, mobile ground, and human- but most appropriate to complete the malicious missions,
packable (i.e., handheld and wearable) platforms according to predicting the response of CUS. On the other hand, the CUS
their mobility and portability levels. Based on the operating can also anticipate the malicious behaviors of mUAVs and
altitude, sky platforms can also be further classified into two establish the effective strategies to defend. In [93], interactive
platforms: low-altitude platforms (LAPs) and high-altitude time-critical situations were studied based on the cumula-
platforms (HAPs). Integrated platforms consisting of ground tive prospect and game theories. Here, an mUAV tries to
and sky platforms that operate both on the ground and sky are minimize the malicious mission completion time, whereas
called hybrid platforms. a CUS platform confronts mUAV to try to maximize the
Each platform can be appropriately employed in a CUS malicious mission completion time of the mUAV. In this
considering their advantages and disadvantages and depend- game, the defense strategies should be carefully designed
ing on the specific requirements of each application. Further- considering the mobility constraint.
more, multiple platforms can be deployed simultaneously and
can cooperate through a network, i.e., a CUS network. The
network should be inter-operable and compatible so that it can 1) STATIC GROUND PLATFORM
coordinate multiple platforms. For example, a static ground The static ground platforms of CUSs constitute the majority
platform equipped with radar, two LAPs equipped with an EO of all platforms (approximately, 54% [38]) and are designed
sensor, and a mobile ground platform providing RF jamming to be deployed on stationary ground facilities, e.g., airports,
can be cooperatively operated as a unified CUS network.2 airfields, nuclear power stations, oil refineries, government
The CUS network can maximize the effectiveness of defense facilities, and households. These platforms are associated
by complementing the limitations of each platform. In addi- with fewer constraints on their size, weight, and power
tion, the CUS network can incorporate any types of platforms, (SWAP). Therefore, static ground platforms are elaborate and
e.g., a hybrid platform that is a specific implementation of the efficient and can be optimized for specific tasks to defend
CUS network. against mUAVs. However, static ground platforms are less
In this section, data-driven insights are discussed for each flexibly able to cope with unpredictable threats from mUAVs.
platform obtained from the current CUSs, consisting of A static ground platform can be equipped with only a
approximately five hundred products, a partial dataset of sensing system (approximately, 43%) or a mitigation system
which is available in the literature [38]. Note that there can be (approx. 25%), or both (approx. 31%), as depicted at the
a dedicated ground platform for C2 systems (i.e., a C2 station top of Fig. 4(a), where the area represents the percentage.
with a human), while this would be difficult for sky platforms. Approximately 60% of sensing systems have a single sen-
Instead, sky platforms, especially HAPs, can equip C2 sys- sor, and 40% of them are equipped with multiple types of
tems without humans or systems to support C2 systems. The sensors, e.g., radar, RF sensors, EO, and IR sensors [94]–
products of CUSs do not include a dedicated system, and [99], as shown in Fig. 4(a)-(i+ii). On the other hand, approx-
imately 34% of mitigation systems have a single mitigator,
2 Throughout the survey in this paper, EO sensors and RF jamming are and 66% of them are equipped with multiple mitigators,
considered as different devices from IR sensors and global navigation satel- such as RF and GNSS jammers [94]–[98], [100], as shown
lite system (GNSS) jamming, respectively. in Fig. 4(a)-(ii+iii).

VOLUME 8, 2020 168679

 H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

FIGURE 3. Platforms of CUS. These platforms are classified into two classes: ground platforms and sky platforms.
Ground platforms are categorized into static, mobile and, human-packable platforms, while the sky platforms are
categorized into low-altitude platforms and high-altitude platforms.

FIGURE 4. Portfolios of the types of ground platforms of CUSs. A partial data set is available in the literature [38]: a) static ground platform, accounting
for approximately 58% of CUSs, b) mobile ground platform, at approximately 16% of CUSs, and c) human-packable platform, accounting for
approximately 26% of CUSs. SS, SM, MS, and MM denote single-sensor and single-mitigator, single-sensor and single-mitigator, multiple-sensor and
single-mitigator, and multiple-sensor and multiple-mitigator, respectively.

It is important to note that integrated platforms equipped (SS) platforms, single-sensor and single-mitigator (SM)
with both sensing and mitigation systems require reliable platforms, and multiple-sensor and single-mitigator (MS)
connectivity and high-level orchestration among the sys- platforms account for approximately 12%, 21%, and 7%,
tems. Thus, static ground platforms are relevant to inte- respectively. The details of these sensors and mitigators are
grated platforms as SWAP constraints are in general absent surveyed in Section V.
compared to mobile and human-packable platforms. Hence,
as shown in Fig. 4(a)-(ii), the platform with multiple-sensors 2) MOBILE GROUND PLATFORM
and multiple-mitigators (MM) accounts for approximately The mobile ground platforms of CUSs, representing approx-
60% of static ground platforms that have both sensing and imately 14% of CUSs [38], are mounted on ground vehicles,
mitigation systems. Here, single-sensor and single-mitigator and they can be agilely deployed to the target location using

168680 VOLUME 8, 2020

H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

the mobility of the vehicles on the ground [101]. Mobile by the eyes of the operators. If a sensing system is employed
ground platforms are suitable for battlefields and dynamically (approximately 19%), it mainly consists of RF sensors [110],
and rapidly changing environments. However, compared to [111]. Only approximately 7% of thuman-packable platforms
static ground platforms, mobile ground platforms have SWAP employ both sensing and mitigation systems [112], [113].
constraints; thus, the available levels and types of sensing and The human-packable platforms equipping multiple sensors
mitigation systems can be limited on this platform. Moreover, [114] take approximately 5% as shown in Fig. 4(c)-(i+ii), and
the utilization of the mobile ground platform is affected by the no MS and MM are employed for the human-packable plat-
capability of the vehicles. forms that have both sensing and mitigation systems as shown
As shown at the top of Fig. 4(b), among all mobile ground in Fig. 4(c)-(ii). Therefore, the human-packable platforms are
platforms, approximately 49% of them have both sensing and relevant as a supplement with other platforms or for a limited
mitigation systems [102]–[104], approximately 25% employ personal purpose.
only a sensing system [105], and remaining 25% have only a
mitigation system [106]. Compared to static ground platforms B. SKY PLATFORM
for which 31% have both sensing and mitigation systems, Sky platforms are systems mounted on certain UAVs, e.g.,
we can infer that an individual mobile ground platform per- airships, balloons, fixed-wing aircrafts, and rotary-wing air
forms as a total solution of an integrated CUS for successful copters. Due to their maneuverability in air, flexible on-
countermeasures, whereas there is a room for a static ground demand placement is possible. Sky platforms are even more
platform to be interoperated with other static ground plat- flexible and more expeditious than mobile ground platforms.
forms without significant SWAP constraints. It is important to note that benefiting from its flexibility,
For the platform with a sensing system, as shown the sky platform can be employed as a multiple-pursuer UAV
in Fig. 4(b)-(i+ii), approximately 56% of sensing systems (pUAV) that tracks and chases mUAVs. In differential game
have a single sensor, while 44% of them are equipped with theory, there have been studies on frameworks to examine
multiple types of sensors, comparable to the static ground pursuit-evasion (PE) problems [115]. By solving the PE prob-
platform. However, as shown in Fig. 4(b)-(ii+iii), nearly half lem, a control scheme can be designed for pursuers to pursue
of the mitigation systems have a single mitigator, while for evaders under position and velocity constraints. To address
the remaining half, the mitigation systems are equipped with PE problems, a linear-quadratic differential game was intro-
multiple types of mitigators [102], [104]–[106]. The ratio of duced in classic work [116], [117]. Multiple players have
the mobile ground platform with multiple types of the mitiga- also been studied [118], where a high-speed pursuer attempts
tors is less than that of the static ground platforms, standing to capture a couple of slow-moving evaders. In other work
at approximately 66%, as the deployment of multiple devices [119]–[121], reach and avoid differential games were pro-
may not be allowed for mobile ground platforms owing to the posed for applications of aircraft control, motion planning,
limited area of the associated vehicles. Furthermore, for the and collision avoidance. Environments in the presence of
same reason, compared to the portion of MM on the static obstacles were also studied [122], while other authors [123]
ground platform, i.e., 60%, the MM portion of the mobile considered a multiple-pursuer and single-evader problem in
ground platform accounts for approximately 32%, as shown which the multiple cooperative pursuers (i.e., pUAVs) capture
in Fig. 4(b)-(ii). a single evader (i.e., mUAV). A single-pursuer and multiple-
evader problem was also studied [124], [125]. In further
3) HUMAN-PACKABLE PLATFORM [126], [127], a scenario in the presence of a defender that
The human-packable platforms for CUSs, accounting for protects an evader against a pursuer was considered. A dis-
approximately 22% of CUSs [38], are designed to be operated tributed algorithm for managing multiple cooperative pUAVs
by an individual by hand. Most human-packable platforms was proposed to mitigate multiple mUAVs [128]. However,
with the sensing systems resemble a backpack or brief- the PE problem is not completely applicable to the design of
case, whereas those with mitigation systems resemble rifles. a CUS. Instead, PE problems can be applied in the case of
Human-packable platforms are lightweight and can be car- pursuers who protect a protective area from evaders [129].
ried by a person, meaning that they are portable. However, Sky platforms are not restricted to traditional missions,
the performance of the human-packable platforms is limited e.g., reconnaissance and attacks, and they recently have been
considering SWAP constraints; e.g., they are associated with rigorously studied for various objectives, such as tracking and
inaccurate detection, tracking, and targeting capabilities, and jamming [130]–[133]. Moreover, recent studies have investi-
also depends on the skill of the operator. Furthermore, due to gated diverse roles of UAVs, for example, as a UAV relay that
the stringent SWAP constraints, most human-packable plat- supports communications between two nodes [25], a UAV
forms only employ mitigation systems (approximately 81%) BS that supports users considering secrecy [78], and a UAV-
without a sensing system, as shown at the top of Fig. 4(c). based edge node that performs computing tasks offloaded by
In these cases, mitigation systems are equipped with multi- nearby users [134].
ple mitigators (approximately 70%, as shown in Fig. 4(c)- On the other hand, sky platforms have critical limitations
(ii+iii)), and the typical mitigators used are RF and GNSS compared to ground platforms. Sky platforms have limited
jammers [107]–[109], while the sensing systems are replaced payloads and battery power such that they can carry only

VOLUME 8, 2020 168681

 H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

lightweight and low-powered sensing systems and/or miti- Typical HAPs are equipped with both sensing systems and
gation systems. Furthermore, sky platforms may generally mitigation systems, where the sensing systems have multi-
require wireless air-to-ground communication links and sys- ple types of sensors, such as EO, IO, and radar types, and
tems, where the communication architecture can be either the widely used mitigator types are projectiles. Surveillance
an ad-hoc network without infrastructure or a centralized HAPs are equipped with only sensing systems. HAPs have
network with a central network node. These requirements been vigorously studied and developed to support other plat-
and the load-and-battery limitations make sky platforms more forms [151]. In such cases, satellite communications can be
challenging compared to ground platforms. considered to link multiple platforms beyond HAPs [67].

1) LOW-ALTITUDE PLATFORM C. CUS NETWORKS

LAPs can fly and hover to cope effectively with mUAVs As surveyed above, each platform has unique benefits; e.g.,
at low altitudes up to a few kilometers [92], [135]. LAPs ground platforms are less constrained by SWAP constraints
are more affordable, and their deployments are quicker and and sky platforms can provide highly flexible on-demand
more flexible than HAPs. Due to the extremely high maneu- deployment and wide operation coverage. On the other hand,
verability and cost-effective mission achievement capability each platform also has certain limitations; e.g., ground plat-
of LAPs, they can play an important role as a part of an forms can support only limited coverage and sky platforms
integrated CUS. LAPs are typically lightweight compared to have stringent SWAP constraints. Therefore, a hybrid plat-
HAPs, and their payloads and fuel/battery power are thus lim- form that consists of ground and sky systems can be con-
ited. To overcome these limitation, energy-efficient designs sidered to offset the shortcomings and enjoy the benefits of
of UAVs has been vigorously studied [131], [133], [136], each system. Furthermore, by leveraging the advantages of
[137]. Moreover, the limited energy/power issue has been ground and sky systems and providing a spatial diversity gain,
tackled through various methods, e.g., the rotation of multiple hybrid platforms can significantly enhance the performance
UAVs, rapid replacement of the batteries, wireless power of CUSs. Hybrid platforms usually have both sensing and
transmission [138], and a tethered UAV whose power can be mitigation systems and consist of various types of sensors as
supplied through a cable [139]. This type of tethered UAV can well as mitigators [144], [152]–[154].
also have a wired communication link for further reliable and An integrated CUS which encompasses hybrid platforms
secure communications [140]. can consist of multiple platforms, such as multiple ground
Most LAPs that engage mUAVs are equipped with only a platforms, sky platforms, hybrid platforms, and combinations
mitigation system only. The typical mitigation method of a of these in a network [154], [155]. The capability of an
LAP is to use either a net or a collision UAV [141], [142]. On integrated CUS is determined by not only the performance of
the other hand, a small percentage of LAPs have a sensing an individual platform but also the properties of the entire sys-
system with most likely a single sensor, i.e., an EO and/or tem of networks. The network can enhance the cooperation
IR sensor [140], [143], [144]. Despite the fact that LAPs among the platforms and thus maximize the effectiveness of
can be equipped with both sensing and mitigation systems, the CUS. Integrated networks are categorized into centralized
their performance is still restricted unless they cooperate with and decentralized networks, as shown in Fig. 5. Decentralized
other types of platforms owing to their limited sensing and networks can be further classified according to the homo-
mitigation capabilities [142], [145]. geneity of the platform [156]. We henceforth introduce two
classified network models and then discuss the appropriate
2) HIGH-ALTITUDE PLATFORM amalgamation of these models.
HAPs fly and hover at high altitudes of up to tens of kilo-
meters [67], [92]. Because HAPs have less stringent SWAP 1) CENTRALIZED NETWORK
conditions, they can be equipped with more systems, such as As shown in Fig. 5(a), a centralized network consists of a sin-
the communication systems and battery/fuel systems. Com- gle high-performance central platform and a cluster of low-
pared to LAPs, HAPs can fly longer and higher and have a performance surrounding platforms. Any type of platform,
wider communication range and the field of vision owing to i.e., ground and sky platforms, can be operated as either the
their high-altitude operability and the high probability of line- central platform or the surrounding platforms. To perform as
of-sight (LoS) environments in communications. Therefore, a centralized C2 system which is a specific implementation
HAPs can effectively counteract mUAVs intruding from high of a centralized network, the high-performance central plat-
altitudes and can also support other platforms. form makes decisions and directs the surrounding platforms
However, HAPs are costly and much more difficult to to neutralize UAVs effectively. Here, the fully centralized
operate compared to LAPs. Moreover, the deployment of network operates effectively when it can obtain access to
HAPs requires more time compared to the time needed to all required information, operate the necessary facilities for
deploy LAPs. Note that traditional aircraft or unmanned making decisions, and disseminate the instructions to the
combat air vehicles developed for reconnaissance and surrounding platforms.
defense/mitigation during military operations can be inter- The centralized network, however, is vulnerable. A break-
preted as high-end HAPs for CUSs [146]–[150]. down or failure of the central platform would affect all

168682 VOLUME 8, 2020

H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

FIGURE 5. Examples of CUS networks: (a) centralized network, (b) decentralized homogeneous network, and (c) decentralized heterogeneous
network.

surrounding platforms, resulting in inefficient CUS opera- ate, the CUS can still operate with slight performance
tion. Furthermore, exchanging information among the plat- degradation. The merit of the decentralized homoge-
forms can cause a long latency because the information must neous network is robustness against malfunctions of the
pass through the central platform. Therefore, robust and inde- platforms. However, each function of the homogenous
pendently dedicated networks are desired to circumvent this platform provides relatively low-quality performance
general concern of centralized network. If there are multiple compared to that of heterogeneous platforms. There-
high-performance platforms, the centralized process can be fore, the platforms may partially cooperate for sensing,
partially distributed. computing, decision making, and neutralizing mUAVs
to maximize the effectiveness of the CUS.
2) DECENTRALIZED NETWORK • Decentralized heterogeneous network: The heteroge-
neous decentralized network consists of multiple types
In a decentralized network, C2 systems are distributed to
of platforms, i.e., heterogeneous platforms, as shown
multiple platforms, as shown in Figs. 5(b) and (c), such that
in Fig. 5(c), where each platform performs only a
each platform in the network cooperatively computes and
few specific tasks, e.g., sensing, computing, decision
makes decisions. A decentralized network can be categorized
making, and neutralization. In this case, each platform
into two models, i.e., a decentralized homogeneous network
should have the capability to execute sufficient perfor-
model and a decentralized heterogeneous network model.
mance for its assigned mission such that any platform
• Decentralized homogeneous network: The decentral- can request that another platform perform a task that it
ized homogeneous network consists of multiple plat- cannot perform. If any platform that undertakes a unique
forms that have identical performance and functions, function fails to complete its role, this partial malfunc-
i.e., homogeneous platforms, as shown in Fig. 5(b). tion may cause a bottleneck and failure of the entire CUS
Thus, unlike the platforms in a decentralized hetero- operation, as the central platform breakdown in a cen-
geneous network, each platform in the decentralized tralized network. However, the well-designed networks
homogeneous network has its own sensing and mitiga- as shown in Fig. 5(c) can resolve this issue. As shown
tion systems. When any of the platforms do not oper- in Fig. 5(c), if any platform does not operate, the other

VOLUME 8, 2020 168683

 H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

two platforms can cooperate to complete the mission. i.e., active sensors, whereas acoustic/ultrasonic sensors
The decentralized heterogeneous network would be a operate passively, i.e., passive sensors. However, sonar
good solution to achieve a tradeoff between robustness typically is used for underwater applications to navigate
and performance. and communicate and is rarely used for UAV detection
owing to the poor propagation characteristics of sonar
IV. ARCHITECTURE waves in air. From this survey, it is revealed that sonar
In this section, the architecture of the integrated CUS is has limited applications, UAV mapping and collision-
introduced. Integrated CUS architectures can be categorized avoidance functions [157]–[160]. On the other hand,
into three types based on their roles, as follows (refer to acoustic/ultrasonic sensors are widely used for UAV
Fig. 6): Sensing systems that gather data from the environment detection; this is discussed further in Section V-A(1).
and transmit the observed data to C2 systems; C2 systems that • Radio wave data: Radio waves consist of waves in
perform computing tasks (e.g., detection/identification and the electromagnetic spectrum, typically in the frequency
tracking/localization algorithms) and make decisions based range from 3 MHz to 300 GHz, and radio wave infor-
on the received data, such as the detection/identification mation has been widely used as UAV detection data.
declaration, localization/tracking declaration, and time and In this case, the wireless channel state information is
method of the neutralization of mUAVs; and mitigation sys- critical to capture radio wave information. For example,
tems that perform mUAV neutralization based on the deci- the path loss is a key metric to determine the presence
sions of the C2 systems. of mUAVs. For detecting UAVs in the sky, it is impor-
Each sensing system, the C2 system, and the mitigation tant to understand air-to-ground (A2G) and air-to-air
system can be equipped in either single or multiple platforms. (A2A) radio channels. A2G and A2A channel models
On the other hand, each platform can employ multiple sys- are different from those of traditional terrestrial channels
tems, i.e., an integrated architecture. However, only a few [75], [76]. Analytic A2G channels can be character-
platform products utilize the integrated type of architecture ized by their LoS and non-LoS (NLoS) components.
because it requires a considerable level of the autonomy to A2G channels are then analyzed according to the LoS
operate the CUS effectively, which could be a burden and probability depending on the environment model [78],
has remained underdeveloped with regard to maximizing [161]. A2A channels tend to have a lower path loss
the performance of CUSs. Hence, most platforms have only exponent than A2G and terrestrial channels [75]. There-
either a sensing or mitigation system and their limitations are fore, exploiting the LoS in A2A channels, sky platforms
compensated by the network among the platforms, as stated equipped with synthetic aperture radar or RF sensors can
in Section III. At this point, the details of each part of the CUS reliably collect radio wave information. To capture radio
architecture are introduced. wave information, radar transmits signals and gathers
the radio data from reflected echo signals, i.e., active
A. SENSING SYSTEMS sensors. On the other hand, an RF sensor collects the
ambient RF signals emitted from mUAVs, i.e., passive
The survey on sensing systems is focused on the information
sensors, as discussed further in Section V-A(2).
collected by sensing systems i.e., the gathering of data, and
how the sensing systems operate. • Light wave data: Compared to radio waves, the light
waves have higher frequencies and shorter wavelengths
with different characteristics. In more detail, light waves
1) GATHERING DATA
include the infrared light (300 GHz–430 THz) and
The sensing systems can collect data such as sound wave visual light (430 THz–750 THz) spectrums. Light
data, radio wave data, and light wave data. Wave data waves have a shorter range than radio waves yet a bet-
can be obtained through various devices, such as sonars, ter resolution owing to the shorter wavelength with a
acoustic/ultrasonic sensors, radar, RF sensors, LiDAR and, higher frequency compared to radio waves. However,
EO/IR sensors. As the details of each sensor are presented in light waves are affected by weather phenomena, such as
Section V, wave data is discussed here. clouds, fog, rain, falling snow, sleet, and direct sunlight,
• Sound wave data: Sound waves are the mechanical due to their short wavelength and have high degree of
waves that include infrasound (up to 20 Hz), acoustic straightness. Hence, the LoS requirements of light waves
(between 20 Hz and 20 kHz), and ultrasound (above are more stringent than those of radio waves. Light wave
20 kHz, up to several gigahertz) waves. Sound waves information in the visual spectrum is intuitive and can
have lower velocities than electromagnetic waves such be analyzed by humans, yet the information collected at
as radio waves and light, and are longitudinal and not dark times, e.g., at night and on cloudy days, is insuffi-
polarizable. Sound waves require a medium (e.g., air cient to provide high-quality visual images. Meanwhile,
and water) through which to propagate. Sound wave infrared radiation is emitted by objects according to the
data can make sensing systems more reliable by pro- black body radiation law. This makes infrared sensors
viding additional data with electromagnetic (EM) wave capable of collecting data such as temperatures irre-
data. To capture sound data, sonar operates actively, spective of the degree of visible illumination. However,

168684 VOLUME 8, 2020

H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

FIGURE 6. Diagram of CUS architecture that consists of sensing systems, C2 systems, and mitigation systems.

because infrared images are detected based on heat greater reliability, (ii) complementary information provided
energy, these images are influenced by the emissivity by sources about different parts of the target, and (iii) coop-
and reflection of sunlight. As active and passive sensors, erative information that is combined into new information
LiDAR and EO/IR sensors are widely used to collect (e.g., multimodal data fusion). The types of data for the
light information, as discussed further in Section V- input and/or output of fusion can be raw data (analog/digital
A(3). signals), features, or decisions. Here, the data type (i.e., data
amount or compression level) which affects the performance
2) DATA FUSION should be carefully designed by considering the tradeoff
The majority of sensing systems have a single type of sensor. between performance and cost, such as the communication
The data collected by a single sensor or identical types of sen- bandwidth and power consumption of the sensors. Further-
sors, however, could be insufficient for accurate and precise more, the processes of data fusion are classified based on
detection/identification and localization/tracking. To offset the JDL and DFIG models into five levels: (i) source prepro-
the limitations of single types of sensors, multiple types can cessing, (ii) object refinements: mUAV classification, identi-
be employed by high-end systems considering the require- fication, and tracking; (iii) high-level inference; (iv) impact
ments and usage environments. Furthermore, instead of sim- assessments: evaluations of threats and predictions; and (v)
ply obtaining results from each sensor type, comprehensive process refinement: resource and sensor management. Fusion
data fusion (i.e., the fusion of sensing information) can be can be performed in a fully centralized architecture, a decen-
implemented [47]. Note that data fusion considers not only tralized architecture, or a distributed architecture according to
multiple sensor types but also multiple identical sensors and the process and fusion capabilities. Note that the majority of
can be implemented in sensing systems and C2 systems. the computation for fusion is performed at a C2 system, which
Data fusion is multidisciplinary in that a clear classifi- can also be centralized, decentralized, and/or distributed.
cation is not established. We introduce four classification Details will be introduced in the next subsection.
criteria to provide a clear understanding of data fusion. Data Some researchers [167] employed a support vector
fusion can be categorized according to the source informa- machine (SVM) with multiple features of sensing data to
tion [162], the data type, the abstraction level [163], joint detect UAVs. The fusion of radar and audio sensors was
directors of laboratories (JDL), or data fusion information studied to identify clearly whether a detected object is an
group (DFIG) models,3 and by the locations at which fusion mUAV or possibly a harmless entity, such as a bird [168].
is performed [166]. The source information can be classified Other authors [169] studied mUAV detection with radar,
as (i) redundant information pertaining to the same target for IR, and EO, as well as acoustic sensors. Multimodal deep
learning was recently studied [170], where data fusion was
3 The process of data fusion, including the data, sensor, and information implemented by extracting multiple features.
fusion steps, is categorized into levels 1 to 4 based on JDL or levels 0 to
5 based on DFIG, where the levels are as follows. Level 0: source prepro-
cessing or subject assessment; Level 1: object assessment; Level 2: situation B. C2 SYSTEMS
assessment; Level 3: impact assessment (or threat refinement); Level 4:
process refinement; and Level 5: user refinement (or cognitive refinement) As mentioned in Section III, the majority of platforms have
[164], [165] only either a sensing or a mitigation system. A central

VOLUME 8, 2020 168685

 H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

platform can perform many roles of a C2 system yet is i.e., operator localization. Generally, operator localiza-
technically discriminated from a C2 system. Some of the tion can be performed only when veiled operators com-
hardware and software of a C2 system can be included in municate with UAVs by RF signals, whereas mUAV
multiple platforms. In other words, C2 system architecture localization can be achieved not only by RF signals
types can be distributed over multiple platforms, and each but also by other data sources. Here, the threat level is
platform can compute and make partial decisions separately. updated according to the localization results.
However, a dedicated C2 system is a core processing unit that Localization for mUAVs should be implemented with-
can orchestrate multiple platforms for high-end performance out GNSS because the GNSS information of mUAVs
of the CUS and can have high computing power. C2 systems is unavailable for CUSs. Localization without GNSS
make decisions about which tasks are required, i.e., orches- (i.e., indoor localization) has been widely studied [173]–
tration, and the threat levels of mUAVs, and perform com- [175]. Indoor localization techniques can be classified
putation for orchestration and decisions. According to i) the into geometric positioning (e.g., triangulation), finger-
distance between the mUAV and the protection area, ii) the printing, proximity analysis, and vision analysis. The
speed and direction of the mUAV, iii) the payload carried by applicable localization techniques for CUSs are geomet-
the mUAV (e.g., explosives), iv) the size and type of UAV, and ric positioning and vision analysis. Geometric position-
v) the attributes of the protective area, the threat level can be ing requires angle and distance information. The angle-
determined, as follows: of-arrival (AoA), received signal strength index, time
• Level.1 (Low): a threat is unlikely. of flight/arrival (ToF/ToA), time difference of arrival
• Level.2 (Moderate): a threat is possible, but not likely. (TDoA) [176]–[178], and round-trip ToF (RToF) [179]
• Level.3 (Substantial): a threat is a strong possibility. can be estimated from sound, radio, and light data, and
• Level.4 (Severe): a threat is highly likely. the estimated information provides source information
• Level.5 (Critical): a threat is expected imminently. for geometric positioning. Estimation by ToF/ToA and
TDoA-based methods may be infeasible for the localiza-
C2 systems make decisions autonomously or by well-timed
tion of uncooperative UAVs, as they require a common
human intervention. Here, human intervention can be a bot-
clock and synchronization. In one study [180], radio-
tleneck to cope with fast-moving UAVs. Therefore, fully
based UAV detection and AoA estimation algorithms
autonomous with the least human intervention possible will
were investigated. In another study [181], AoA estima-
enhance the performance of CUSs.
tion techniques using a directional antenna array were
proposed to localize UAVs. A visual analysis can also
1) ORCHESTRATION be employed for localization. The visual analysis is
The orchestration procedure of C2 systems in an integrated implemented based on light information (i.e., captured
CUS can be divided into five steps, as follows [49], [50]. Note images). The obtained information is discriminated with
that the threat level can be updated during every step, and step irrelevant background (e.g., buildings and static objects)
(v) can be directly executed while omitting the other steps to estimate the positions of mUAVs [173], [182]–[184].
depending on the threat level. However, a depth camera is needed to estimate the dis-
(i) Detection/identification: To detect any suspicious tance between an mUAV and a sensor. The distance can
object, C2 systems initially gather data from the sensing also be estimated with prior knowledge of the mUAV
systems. C2 systems may then perform data fusion by without a depth camera [185].
dividing the tasks of extracting features and making (iv) Tracking: According to the updated threat level,
decisions (i.e., identification) with sensors as to whether the C2 systems determine whether to track the detected
the detected object is a UAV or another small object, e.g., mUAV. A sky platform is an effective tool capable of
a bird, kite, or balloon. The decision can be made from physically tracking an mUAV, i.e., chasing it. On the
raw data, feature data, or local decision data. Further- other hand, tracking can be interpreted as the algorith-
more, C2 systems can classify the payload carried by mic tracking of the target UAV by C2 systems and
the UAV to determine the threat level [171], [172]. sensing systems. Tracking can also be implemented
(ii) Authorization: When C2 systems conclude that a by data fusions such as localization. For algorithmic
detected object is a UAV, they can then verify whether tracking, an extended Kalman filter, a particle filter,
the detected UAV is authorized or unauthorized. Accord- and template matching are widely employed for gen-
ing to the decision with regard to verification of autho- eral tracking from ground sensors [130], [183], [186].
rization, C2 systems update the level of the UAV threat. Note that authorized UAVs can actually be camou-
(iii) Localization: If the threat level exceeds a predefined flaged or stolen/spoofed/hacked by malicious operators,
level (e.g., level 2), C2 systems identify where the and UAVs can veil their intentions and pretend to be
detected UAV is located and/or whether it is heading authorized until the moment they present the harmful
toward a sensitive protecting area, i.e., mUAV local- threat. Authorized UAVs can operate in a malicious
ization. Localization for the operators of mUAVs can manner abruptly. Thus, C2 systems must continue to
be performed to investigate and prevent future threats, observe/track even authorized UAVs. While tracking an

168686 VOLUME 8, 2020

H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

mUAV, the threat level must be updated according to the C. MITIGATION SYSTEMS
tracking result. According to the threat level as determined by the C2 system
(v) Decision on Neutralization: C2 systems can make and following the regulations of relevant authorities, several
decisions to neutralize UAVs from the updated threat mitigation systems can be simultaneously activated and coop-
level. Neutralization methods include controlling, erate to mitigate mUAVs effectively. Based on the strength
warning, disrupting, disabling, and destroying [187]. of the threat level and countermeasures against mUAVs,
Following the regulations of the authorities and accord- mitigation systems can warn, control, disrupt, disable, and
ing to the neutralization strategy, the neutralization destroy by utilizing various mitigators, such as RF/GNSS
method is determined by the C2 system. To increase jamming, spoofing, high-power microwaves (HPMs), lasers,
the effectiveness, multiple mitigation systems with nets, eagles, projectiles, and collision UAVs [187].
various neutralization methods can be operated (i) Warning: With knowledge of the utilized communica-
simultaneously. tion system of the mUAV, mitigation systems can warn
and neutralize the mUAV by communicating with the
operator of the mUAV on restrained terms when the
2) COMPUTATION
threat level is Level 2(Moderate). Because the pilot of
Throughout the integrated CUS procedures, high computing the mUAV can sabotage the UAV, which could be a
power is required to improve the accuracy and effectiveness danger for civilians if the mUAV is flying or hovering
of detection/identification, localization, tracking, and neu- over habitations, the warning would be the first neu-
tralization. Outstanding computing performance is required tralization4 strategy before other mitigation methods are
to implement state-of-the-art data fusion schemes, detection- used. To the end, mitigation systems should include a
localization-tracking algorithms, and for orchestral multi mit- communication system and the capability to provide the
igation system operation. The computing complexity can direction of the flight such that the mUAV can deviate
increase exponentially for integrated CUSs that require the from the unauthorized route to avoid an intrusion.
capability to cope with multiple UAVs and state-of-the-art (ii) Control: Instead of warning the mUAV operator, direct
algorithms. To this end, C2 systems must provide high com- control of the mUAV can be implemented via spoofing.
puting power. This requires more sophisticated and high-end tech-
Meanwhile, centralized computing can be a bottleneck in niques and devices when the threat level is higher
CUSs. A breakdown or/and failure of centralized computing than or equal to Level 3(Substantial). By taking control
can limit the system’s ability to protect the skies. Decentral- of the mUAV, mitigation systems can land the mUAVs
ized or distributed computing can provide a robust network safely and immediately on the ground. If there is a return
without system bottlenecks, while a single C2 system on a to home (RTH) mode in the mUAV, the RTH mode
global platform can lead to a vulnerable network. Decentral- can be activated [190]. However, owing to the lack of
ized/distributed computing can be implemented on platforms standards, protocols, and regulations, it is difficult to
with computing capabilities cooperatively sharing computing implement control methods practically.
tasks. (iii) Disruption: Disruption refers to interrupting the opera-
Recently, cloud computing, where a cloud with powerful tion of an mUAV. Mitigation systems can disrupt poten-
computing capabilities performs highly complex tasks, has tial mUAVs that can threaten a protected area when the
emerged. The cloud can provide high computing power and threat level is higher than or equal to Level 4(Severe).
network management given its benefits of vast resources. Typical disruption methods are cyber attacks, such as
However, cloud computing is centralized and has the draw- jamming and spoofing. By using jamming and spoofing
back of latency. Fog computing or edge computing has methods, mitigation systems disrupt the mUAV so that it
also emerged to deal with this problem. Fog computing cannot be operated with full maneuverability. Once the
can cope with latency-sensitive applications using network mUAV is disconnected from the operator by disruption,
edge nodes. Network edge servers (or cloudlets) with a the RTH mode can be activated [190].
distance closer than the cloud compute tasks and there- (iv) Disabling: Compared to disruption, which causes UAVs
fore decrease the propagation delay. On the other hand, to malfunction, disabling UAVs is harsher when the
the cloud can be reached by passing several networks on threat level is higher than or equal to Level 4(Severe).
which network managing operations (e.g., routing, medium Strongger RF/GNSS jamming, spoofing, and HPMs
access control) are needed. However, the computing latency can disable mUAV operation in a non-physical manner.
of fog computing is greater than that in cloud comput- In addition, a net catcher or eagles, which are the kinetic
ing. Therefore, task offloading must be rigorously designed mitigation systems, can disable UAVs physically.
based on this tradeoff. Note that employing the sky plat- (v) Destruction: Destroying mUAV is the harshest means
form (not only the ground platform) as a cloudlet has also of physically neutralizing an mUAV by using weapons
been vigorously studied [134], [188]. Readers can refer to
one earlier study [189] and the references therein for more 4 Neutralization and mitigation are interchangeably used throughout the
details. paper.

VOLUME 8, 2020 168687

 H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

such as lasers, projectiles, and collision UAVs [191]. of audible (between 20 Hz and 20 kHz) and ultrasound
These system can be activated when the detected mUAV (above 20 kHz, up to several gigahertz) waves. Most UAVs
is too close to a secure-sensitive area, such as an air- generate sound from the engines/motors and/or rotors. Mini-
port, airfield, nuclear power station, oil refinery, public UAVs generate buzzing and hissing sounds in the frequency
infrastructure, government facility, or military facility range of 400 Hz to 8 kHz [192], which can be detected by
and related areas, and/or then they are too fast to ver- the acoustic sensors. The gathered sound data can be com-
ify the threat or to use other more moderate neutral- pared to libraries of acoustic signatures to discriminate UAVs
ization methods. In urgent situations, i.e., threat Level from other, similar objects [192]. For example, DroneShield
5(Critical), the multiple destroying systems can be acti- built a database of the acoustic signatures of various UAV
vated and cooperate to improve the protection capability. models to prevent false alarms due to ambient noise [193].
The destruction of an mUAV may have a knock-on Alsok’s detection system employs acoustic sensors to detect
effect from the debris of the mUAV and the explosion. the rotating propellers of UAVs and compare the detected data
Thus, in urban environments where many people can with the acoustic signatures in a database [194]. However,
be injured, mitigation systems need to determine the the libraries of acoustic signatures do not cover all types of
destruction time. The destruction can also be deferred proliferating UAVs over various fields. Furthermore, acoustic
to locate and capture the operators of mUAVs. Note that sensors cover a limited range, and acoustic/ultrasonic data is
the physical destruction can be a last resort for mitigating vulnerable to wind and surrounding ambient noise sources.
mUAVs. Though the LoS environment can enhance the detection per-
To overcome the limitation of single mitigator/UAV of formance, it is not required for acoustic/ultrasonic sensing.
CUSs, the operation of the multiple UAVs is desired. To this To overcome the limitations of acoustic/ultrasonic sensing
end, the newest wireless communication technologies capa- and to bolster the performance capabilities of other types of
ble of supporting/controlling numerous devices and with sensors, various techniques and algorithms have been studied
ultra-reliable and low-latency communications, e.g., 5G and using acoustic/ultrasonic data.
B5G, are recommended for CUSs. A microphone can be used to detect a UAV [195].
To increase the detection range, the arrays of microphones
can be used [196]. Localization and tracking of UAVs were
V. CUS DEVICES AND FUNCTIONS
studied with an acoustic array using calibration and beam-
Sensors and mitigators are essential components that com-
forming [197]. In another study [198], a classifier with two
pose the sensing and mitigation systems, respectively,
layers was proposed, where the first layer determined the
as introduced in Section III and as shown in Fig. 6. Each
existence of a UAV and the second layer determined the
sensor and mitigator has unique characteristics, limitations,
UAV type, e.g., fixed-wing or rotary-wing. Machine learning-
and shapes, as shown in Fig. 7. Additionally, multiple sensors
based algorithms such as the SVM and k-nearest neighbor
and mitigators can be deployed on a single platform, as shown
(k-NN) algorithms, as well as neural networks were studied
in Fig. 7. In this section, we introduce the details of sensors
to classify the time- or frequency-domain acoustic/ultrasonic
and mitigators and their functions.
signals generated from UAVs [199], [200].
A. SENSORS 2) RF SENSORS
A sensor is generally a device, module, machine, or subsys- RF sensors capture ambient EM signals emitted from
tem that detects and reports events or changes of a monitored mUAVs or remote operators to detect mUAVs. The major-
surrounding environment. Herein, the sensors of sensing sys- ity of commercial UAVs are remotely controlled by their
tems for CUSs are inteded to detect and report UAVs and operators. For example, UAVs and operators communicate
can be classified as active or passive sensors. Active sen- telecommand and telemetry information, such as altitude,
sors, such as radar and LiDAR, transmit waves and receive position, battery life, and video data. Hence, RF sensors
reflected waves to collect data. On the other hand, passive can detect mUAVs unless the mUAV is preprogrammed and
sensors such as the acoustic/ultrasonic sensors, RF sensors, autonomous. Because RF sensors are easy to implement and
and EO/IR sensors, receive ambient waves which are emitted have low computational complexity, they have been studied
from UAVs. Sensors can be also categorized according to the for various systems. In one such study [201], the average sig-
frequency of the transmitting and/or receiving waves. Herein, nal strength measured by several RF sensor nodes was used
sensors are surveyed based on the wave frequencies, from to detect UAVs. Using Wi-Fi receivers and software-defined
low to high, as shown in Fig. 6. They are also categorized radio boards, RF sensors eavesdrop on the link between the
in Table 6. mUAV and the controller and capture the vibrating patterns of
the UAV body for UAV detection [202], [203]. The detection
1) ACOUSTIC/ULTRASONIC SENSORS and classification of micro UAVs from the RF signals were
Microphones are pressure transducers that convert sound also studied based on machine learning approaches [204].
waves into electrical signals and are thus widely used as RF sensors are widely applied to various systems owing to
the acoustic/ultrasonic sensors that detect the spectral range their simplicity, yet they have several limitations. RF sensors

168688 VOLUME 8, 2020

H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

FIGURE 7. Sensors and mitigators. Note that radar, RF sensor, jamming, spoofing, and high-power EM employ antennas and their functions
can be implemented with the same hardware; therefore, their appearances are similar to one another. The platforms show sensing and
mitigation systems equipped in ground static platforms, ground mobile platforms, human-packable platforms, LAP, and HAP.

TABLE 6. Characteristics and Limitations of Sensors.

have poor target detection reliability and high false alarm information [205]. Signals sharing the same frequency band
probability rates. Because the RF sensor is passive, it does as UAVs, i.e., electromagnetic interference, make RF-based
not provide the range information of the mUAV. Knowledge UAV detection more challenging. Furthermore, identifying
of the spectrum band in use is required for detection. Further- MAC addresses is only possible for disclosed-to-the-public
more, knowledge of modulation protocols, e.g., the frequency MAC addresses.
hopping spread spectrum, the direct sequence spread spec-
trum, and orthogonal frequency division multiplexing, and/or 3) RADAR
the identification of media access control (MAC) addresses is To determine the range, angle, or velocity of an mUAV, radar
required to improve the fidelity of the detection performance is widely used as an active sensor in sensing systems in a
[48]. Here, spectrum sensing can be employed to acquire CUS. A radar system consists of a transmitter, a receiver, and

VOLUME 8, 2020 168689

 H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

a processor [211]. The transmitter radiates EM signals whose wavelengths (300 GHz–430 THz) reflected from mUAVs to
frequency ranges typically between 3 MHz and 300 GHz detect them under daylight conditions. On the other hand, IR
depending on the application. The EM signals are reflected sensors, i.e., thermal cameras, detect the infrared spectrum
by the mUAV and return to the radar. The returning EM to capture the heat signature (resolutions as low as 0.01◦ C )
signals reflected from the mUAV provide essential informa- radiated from mUAVs and thus can detect targets even with-
tion with which to obtain the mUAVs’ location and speed. out sufficient light, e.g., during the night and cloudy and/or
Thus, the amount of the received signal power is critical to dark days. The spectrum should be determined based on the
determine the detection performance of the radar. However, expected temperature of the target object. An IR sensor can
because the reflected radar signals captured by the receiving detect heat emitted from the motors and engines of mUAVs
antenna are very weak, they need to be amplified at the pro- [45], [231], [232], where IR cameras with shorter wavelength
cessor. The reflected radar signals captured by the receiving provide better performance to capture fast-moving bright and
antenna are inversely proportional to the frequency, whereas small targets than long-wavelength IR cameras [45].
they are proportional to the radar cross-section (RCS), a mea- Passive EO/IR sensors provide only two-dimensional
sure of how detectable an object is that depends on the (2D) images. Accordingly, to enhance the detection perfor-
material, size, and location (i.e., the distance and incident mance, various machine learning- and deep learning-based
and reflected angles) of the mUAV. From the reflected radar approaches have recently been employed. Machine learning-
signals, the processor can calculate the round-trip time (i.e., based approaches, e.g., SVM and k-NN, classify objects
ToA) and the frequency shift due to the Doppler effect to based on predetermined features, whereas deep learning-
estimate the distance and velocity information of the mUAV. based approaches are typically convolutional neural networks
However, traditional radar systems are designed to detect (CNN) without specified features. For example, the use
legacy (manned) aircraft with high velocities and a large RCS, of neural networks has been rigorously investigated and
and they are inappropriate to detect slow-moving and low- developed for EO/IR sensors [130], [233]–[239]. In [233],
flying mUAVs with a small RCS [206], [224], [226], [227]. a regression-based approach was studied for its ability to
To circumvent this issue, the micro-motions of vibrating classify and detect UAVs. In that case, the training dataset
(by engines or motors) and rotating (by propellers) struc- was also provided. The training data set can be artificially
tures of UAVs [218]–[220], which cause a unique micro- generated for the CNN [240]. Various neural networks, such
Doppler signature (MDS), have recently been used for radar as that by Zeiler and Fergus of the Visual Geometry Group
detection. Research has shown that quadcopters, hexacopters, and another entitled ‘You Only Look Once’, have also been
and octocopters have different MDS characteristics [222], assessed for UAV detection [234], [240]. Robust algorithms
[223]. Radar can detect mUAVs by analyzing the MDSs of for static/moving cameras were designed to propose candi-
mUAVs [206], [215]–[217]. The joint time-frequency anal- date regions and classify UAVs with birds [238], and the
ysis method, e.g., short-time Fourier transform, can also be onboard UAV system was devised to detect and chase other
utilized to analyze the radar MDSs of UAVs [221]. UAVs using a lightweight camera and a low-power algo-
Various types of radar to detect objects with small RCSs rithm without a GNSS service [237]. A pUAV detecting a
have been studied. An unmodulated continuous wave (CW) target mUAV based on template-matching algorithms with
Doppler radar system with a long dwell time can capture a morphological filter was also considered [130]. In another
rich information to deal with small UAVs with a small RCS study [239], an onboard UAV-Net detector was proposed to
[207]–[210], though it cannot obtain the target range [211]. detect small objects. Algorithms based on CNNs and spatio-
A frequency-modulated CW radar system can estimate the temporal filtering have also been proposed to detect and track
ranges as well as velocities of multiple targets simultaneously mUAVs and discriminate mUAVs from birds [235]. In other
[46], [212]–[214]. On the other hand, ultra-wideband (UWB) work [236], a super-resolution object-detection method for
radar generates an extremely narrow pulse, resulting in wide- detecting UAVs was designed.
band utilization. UWB radar can be employed for high- Though EO/IR sensors have been widely studied and
resolution ranging resulting in an accurate ToA. Experimental utilized for object detection, they have several limitations.
results show that MDSs induced by mini-UAVs and birds are The detection performance capabilities of EO/IR sensors are
significantly different, and it was found that mini-UAVs and highly degraded under NLoS environments. Further, good
birds can be distinguished based on features caused by the focusing capability and multiple cameras are required for
flapping wings of the birds and the unique MDS [206], [228]– EO/IR sensors to perform multi-direction detection. EO/IR
[230], [243]. It was also found that millimeter-wave radar sensors are susceptible to adverse weather conditions and
can provide high-fidelity micro-Doppler echoes from a mini- may fail to detect objects near the horizon. A mUAV with
UAV from the very rapidly rotating propellers [224], [243]. temperature comparable to background objects may be chal-
lenging for IR sensors to detect [45].
4) EO/IR SENSORS
EO sensors detect EM waves that range from the infrared 5) LiDAR
(300 GHz–30 THz) up to the ultraviolet (larger than Similar to radar, LiDAR detects mUAVs from signals return-
790 THz) frequencies. Typically, EO sensors capture visible ing after reflecting off of the mUAVs. Contrary to radar,

168690 VOLUME 8, 2020

H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

TABLE 7. Characteristics and Limitations of Mitigators.

LiDAR emits laser light (typically, 300 THz–500 THz) to Because nonphysical mitigators use EM waves, instanta-
measure the range information from the mUAV. LiDAR can neous maneuvers are possible and are not affected by certain
provide 3D representations using differences in return times. aspects of the physical environments, such as gravity and
Therefore, LiDAR can differentiate a target object from a wind. Thus, nonphysical mitigators can readily aim at target
complex background [242]. mUAVs. Nonphysical mitigators can be realized by vari-
In one study [241], the authors proposed an algorithm ous methods, such as high-power electromagnetics, lasers,
for detecting small UAVs and generating 3D coordinates by and cyber-attacks (e.g., RF/GNSS jamming and spoofing,
employing LiDAR. In another [101], detecting small UAVs deauthentication attacks, zero-day vulnerabilities, cross layer
was assessed using a LiDAR system mounted on a vehicle. attack, multi-protocol attack, denial-of-service on UAV/GCS,
However, LiDAR has a short range to detect objects and address resolution protocol cache poisoning [286]). In our
requires a LoS environment owing to the high frequency survey, among the cyber-attacks, we focus on RF/GNSS
of the laser light and its low energy. To extend the range jamming and spoofing which are the majority of the cyber-
of detection, a data augmentation method and a detection attacks to mUAVs. See [51], [249], [286], [287] and refer-
algorithm were studied for detecting UAVs using LiDAR ences therein for the comprehensive survey of cyber-attacks.
[242]. Moreover, as mentioned in Section IV-A1(1), LiDAR
is affected by weather phenomena, such as clouds, fog, rain, • RF/GNSS jamming: The RF jammers can dis-
falling snow, sleet, and direct sunlight. rupt or disable mUAVs by interfering with their commu-
nication links. By interfering with the communication
between mUAVs and the malicious operators, jamming
B. MITIGATORS
decreases the signal-to-noise ratio (SNR) of the mUAV
Mitigation methods have been categorized into nonphysical
and disrupts the mUAV [133]. To recover the disrupted
and physical methods based on whether there is physical
communications, the communication signal between the
damage to the mUAV, as summarized in Table 7 and shown
mUAV and the malicious operators must increase, which
in Fig. 7. In this section, nonphysical and physical mitigators
exposes them clearly to the mitigators. Once the commu-
are surveyed.
nication link is jammed and degraded, the mUAVs may
lose the remote control link and may descend or initiate
1) NONPHYSICAL MITIGATORS a RTH mode.
Nonphysical mitigators employ EM waves to disrupt, dis- There are several jamming schemes. A jammer can
able, and/or destroy mUAVs. Nonphysical mitigators per- transmit all of its power on a single frequency (spot
form the invisible, silent, and mild mitigation, as there is jamming), shift the power rapidly from one fre-
no physical contact between the mitigator and the mUAV. quency to another (sweep jamming), or transmit power

VOLUME 8, 2020 168691

 H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

simultaneously over a range of frequencies (barrage including a navigation system and a communication
jamming). In addition, jammers can be classified as system [249], various vulnerabilities of mUAVs can
active jammers and reactive jammers. The active jam- be considered to increase the capabilities of spoofing.
mer transmits RF signals continually, or does so ran- With rigorous analysis and overwhelming technologies,
domly to save energy. The deceptive jammer, a type spoofing by attacking the vulnerabilities of operating
of active jammer, causes the UAV to receive pack- systems, GNSS systems, and wireless communication
ets continuously without a gap such that the mUAV links can be implemented.
remains in a receive mode. The reactive jammer trans- • High-power electromanetics: A high-power EM wave
mits signals only when it detects that the monitored can disable an mUAV by impairing its electronic sys-
spectrums/channels are occupied by unknown signals, tems, and these methods can be categorized into two
i.e., mUAVs, [244], [245]. However, RF jamming can be classes: those that use narrowband waves and those that
ineffective for autonomous mUAVs that do not require use wideband waves. Narrowband EM waves include
any remote control or for mUAVs that follow a pre- high power on a nearly single-tone frequency. A high
programmed route via global positioning system (GPS) power narrowband EM wave is referred to as HPM.
checkpoints [288]. Thus, GNSS jamming is required to HPM can couple with the UAV and cause damage such
compensate for the limits of RF jamming. that it becomes disabled. HPM requires very high power,
GNSS jammers interfere with navigation systems. i.e., on the order of thousands of volts on a single fre-
Because the GPS signal comes from a satellite, its power quency [253]. The directed energy of HPM can be used
is weak and vulnerable to jamming signals. Once the to crash a UAV [254]. Finding an effective frequency to
mUAV loses the GNSS signal, it will hover or land cause malfunctions in mUAVs is the key issue.
without completing its mission [247]. However, GNSS On the other hand, the wideband EM wave has short
jamming can be ineffective for mUAVs equipped with pulses in the time domain. The energy is distributed
inertial measurement unit (IMU) sensors and encrypted over a wide band, and the wideband wave EM has a
signals for the navigation. Therefore, the compensation low energy density over the bandwidth. Note that a non-
between RF and GNSS is required. nuclear EMP can hardly be implemented with a large
It is important to note that sky platforms can be low-inductance capacitor that is discharged into a single
employed for effective jamming mitigators, as the jam- loop antenna.
ming performance can be dramatically improved as However, high-power EM waves should be precisely
the distance between the mitigators and the mUAVs directed toward the target mUAV to effectively mitigate
becomes shorter [49], [132], [246]. it; otherwise, the lethality is significantly decreased,
• Spoofing: Given the overwhelming technology and/or i.e., some devices, e.g., radar and RF sensors, can still
knowledge of mUAVs, taking control of mUAVs or com- operate partially after this type of mitigation [255]. Here,
manding mUAVs to detour away from a protected area the issue is that it is difficult to evaluate the kill assess-
is possible, a technique also known as spoofing. Spoof- ment after mitigation.
ing mitigators can disrupt, disable, or take control of • Lasers: While lasers can be employed as laser range
mUAVs. Spoofing mitigators for mUAVs counterfeit finders and designators, laser as mitigators can dis-
RF or GNSS signals to neutralize mUAVs. Advanced able or destroy mUAVs with directed energy [256]–
technologies which determine fully the communication [263]. An electrolaser ionizes the path to the UAV and
protocol stacks, GNSS services, and vulnerabilities of emits an electric current down the conducting track of
the mUAVs are required to implement spoofing. ionized plasma. Lasers can be categorized into low-
GNSS spoofing is a common method when the pro- power lasers and high-power lasers [255]. Low-power
tocols (e.g., code and modulation types) are known. lasers can neutralize (dazzle) the sensitive EO/IR sen-
GPS spoofing can cause mUAVs to hover, engage the sors of mUAVs. High-power lasers that operate at the
autopilot, land, and misdirect to the spoofed route [247], mega-watt level can burn a hole in the mUAV and
[248]. Appropriate spoofing strategies are needed for destroy it. Laser mitigators are affordable compared
different types of mUAVs to manage them when they to physical projectiles [289]. However, laser mitigators
lose their lock on the authorized GNSS signals [247]. require challenging research and development and are
The spoofing of remote control signals can also be sensitive to adverse weather conditions. Furthermore,
implemented by analyzing the communication proto- high-power lasers require accurate directions and suffi-
cols in use [249], [250]. Taking full control of mUAVs cient time to track the mUAVs.
is possible [250], [251] if the protocols are known
and available at the mitigators. Vulnerabilities of Wi- 2) PHYSICAL MITIGATORS
Fi-based UAVs have been studied [249]. Cellular- Contrary to nonphysical mitigators, physical mitigators dis-
connected UAVs [252] can also be spoofed by analyzing able and destroy mUAVs physically. Physical mitigators are
the vulnerabilities of cellular networks. Furthermore, effective, and the results of whether the neutralization was
because mUAVs consist of various embedded systems successful are obvious. Physical mitigators require accu-

168692 VOLUME 8, 2020

H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

rate aiming and/or tracking of the mUAVs to remain phys- in various areas. However, eagle mitigators can easily
ically close to the mUAV to effectively neutralize. Physi- be injured by the blades and propellers of mUAVs, and
cal mitigators can employ projectiles, collision UAVs, nets, their use is limited to slower and smaller mUAVs relative
and eagles. to the speed and size of the eagles. Furthermore, eagle
mitigators may not be appropriate to mitigate multiple
• Projectiles: Mitigators that employ projectiles can mUAVs simultaneously.
destroy mUAVs. Projectiles include machine guns,
munitions, guided missiles, artillery, mortars, and rock- VI. CUS MARKET
ets. Guided projectiles require a guidance system to Skyrocketing growth in the UAV industry has created both
track and hit the mUAVs. In July of 2014, Israel used positive and negative externalities. Well-meaning users of
a Patriot missile to shoot down an incoming reconnais- UAVs have successfully benefitted from diverse UAV appli-
sance UAV from Gaza [264]. In 2019, SmartRounds cations that range from recreation to emergency rescue appli-
Inc. announced a 40 mm missile system for anti-UAV cations, as introduced in Section II-A. Malevolent users of
munitions, which can be deployed on ground and sky UAVs also have quickly caught up with the possible malicious
platforms and that operate at high velocity [265]. The applications of UAVs, such as terror, security breaches, and
projectile is equipped with a vision sensor for object invasions of privacy, to name a few. As a result, market needs
detection and tracking. However, precise aiming consid- for counteracting the negative externalities of UAVs have
ering gravity and wind is required, and the cost of the skyrocketed, in tandem with the recent growth of the civil-
projectiles per shot is high. Furthermore, the mUAVs go ian UAV industry. However, these market needs are bound
out of control and crash to the ground, possibly causing to be multi-faceted due to the distinctive characteristics of
collateral damage. the CUS market, such as its complete dependence on the
• Collision UAVs: Collision UAVs with detection and UAV industry or the possibility of cannibalizing existing
tracking capabilities can follow the mUAVs to crash markets. In this section, the landscape of the CUS market is
into and destroy them. A collision UAV requires high scanned to identify market patterns and anomalies. The CUS
speed to pursue the mUAV, e.g., 350 km/h [266], and market is analyzed in terms of rivalries and major acquisi-
is effective for contiguous small mUAVs in protected tions/partnerships among incumbents, suppliers or partners
areas. Collision UAVs can employ a computer-vision- of incumbents, as well as those complementary to incumbents
aided object-detection method and carry explosives to and emerging organizations entering the CUS market. Based
maximize the collision impact [268]. More examples on the market analysis, the distinct characteristics of the
of collision UAVs can be found in the literature [191], CUS market are identified and elaborated. Lastly, practical
[269]–[271]. Collision UAVs can be interpreted as a implications for industry practitioners, especially those which
hybrid consisting of a missile and a small UAV. Colli- consider entering the CUS market, such as telecom service
sion UAVs are disposable and cause collateral damage, providers, are also identified.
similar to projectiles. However, collision UAVs require
a relatively large neutralization delay compared to pro- A. DANCING LANDSCAPE OF THE GLOBAL CUS MARKET:
jectiles. WHAT DOES IT LOOK LIKE AND HOW IS IT CHANGING?
• Nets: Net catchers ensnare and demobilize mUAVs. The The size of the commercial UAV industry is anticipated to
net can be projected by a net cannon [272]–[276] or can make an upsurge, reaching USD 6.3 billion in 2026 with
be carried by sky platforms [49], [277]–[283]. Nets a compound annual growth rate (CAGR) of 23.37% from
can be a solution to mitigate small mUAVs which the market size of USD 1.2 billion in 2018 [290]. The
are difficult to neutralize by guns or guided missiles prospects of the commercial UAV industry started to become
[282], [283]. In one study [284], a portable mitigator especially hopeful when Jeff Bezos, the CEO of Amazon,
was demonstrated to be able to capture UAVs. Nets made a surprising appearance on ‘‘60 Minutes’’ in Decem-
can be equipped with parachutes to ensure that the ber of 2013, and announced Amazon’s future plan to launch
UAV descends safely for forensic analysis and to pre- a drone delivery service called Amazon Prime Air. Around
vent collateral damage to other facilities [49]. However, the same time, however, the portents of commercial UAVs
the effective range of net mitigation is short. going rogue had been progressively noticeable. A rogue
• Eagles: For centuries, people of the Altai region have drone was spotted at Gatineau jail in Quebec in Novem-
trained themselves in the art of eagle hunting. They have ber of 2013, which obviously was an attempted contraband
trained eagles to catch small animals. Motivated by the drop-off. A few days later, guards at Georgia State Prison
people of Altai, Dutch and Scottish police trained eagles spotted a six-rotor drone carrying packs of tobacco hovering
as a mitigator of CUSs to neutralize and catch mini- over the prison compound. Six and a half years later, while
UAVs [49], [285]. Eagle training does not require high we are still waiting for Amazon drones to drop off packages
technology. To train and breed a mitigator eagle may onto our door steps, the potential threats of mUAVs have
require fewer human resources than other mitigation surged, subsequently intensifying the market need for CUS
devices that are developed by researchers and engineers solutions.

VOLUME 8, 2020 168693

 H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

1) MARKET SIZE AND GROWTH

The CUS market remained in its embryonic stage throughout
the 2010s only to witness ‘hockey stick’ growth recently.
The current CUS market is estimated to have reached the
size of USD 1 billion [291] and is expected to grow to USD
4.5 billion by 2026 [292]. The five-year forecast for the CUS
market growth ranges, depending on market research firm,
from a CAGR of 16.8% [293], 37.2% [294], to 41.1% [291].
Fig. 8 shows the current size of the CUS market and its
anticipated growth over the next five years as estimated by
Drone Industry Insights, the German market research and FIGURE 8. CUS (counter-drone) market size and forecast 2019–2024 [295].

analysis company [295].

The most critical and direct driver is the proliferation of
2) GEOGRAPHIC COMPOSITION low-price UAVs that already have created a mass market.
The geographic distribution of the CUS market highlights the While regulations have been keeping pace with the growing
dominance of North America, whose market share accounts UAV industry, as stated in Section II-B, market demands and
for more than half of the global CUS market [293], [296]. regulatory changes do not move in sync, forcing regulatory
North America’s CUS market dominance is mostly due to bodies to make continuous updates. Conflicting perspectives
the prolonged and extensive R&D investment by the US on fundamental regulatory issues, such as categorizing UAVs
Department of Defense (DoD), especially up to 2016, and the as either ‘flying objects’ or airplanes, prohibits regulatory
subsequent procurement of CUS solutions. Starting in 2016, bodies from reaching a consensus with regard to safety
the DoD shifted its investment focus from R&D to integrating levels. For instance, the National Aeronautics and Space
existing technologies and solutions into more comprehensive Administration (NASA) considers a relatively narrow indus-
programs [297]. In order to drive this shift further, the DoD try of UAVs when categorizing them as something between
requested USD 500 million for CUS development for the road and air devices, whereas the European Aviation Safety
2020 fiscal year. In sum, although the underlying mechanism Agency (EASA) maintains a broader view of UAVs, catego-
has evolved from system development to system integra- rizing them at the level of an airline [300]. The expanding
tions, North America’s CUS market dominance is expected UAV mass market poses both malicious and benign threats
to remain strong for the foreseeable future. Europe is the by providing malevolent actors with an asymmetric capability
next most active geographic region for the CUS market, to launch attacks and enabling recreational users to cause
where the growth rate is estimated to remain steady [298]. unintended security breaches. The other direct driver is the
The strongest driver of the European CUS market’s medium emerging market requirements specifically for portable CUS
yet steady growth is the presence of globally renowned tra- solutions, i.e., human-packable and mobile platforms.
ditional defense corporations such as the Thales Group in Indirect drivers are in action as well. Coincident with
France, Saab AB in Sweden, and BSS Holland BV in the public exposure to drone swarm technology spiked through
Netherlands. Multiple market research and analysis organi- international events such as the Winter Olympics [301], major
zations collectively identify the Asia Pacific CUS market as news outlets, e.g., the Financial Times, have recently started
possessing the most substantial growth potential in the near to warn the general public about the potential threat of
future [292], [298], [299]. Rapidly increasing government drone swarms [302]. Public perceptions that drone swarms
expenditures on defense infrastructure, particularly that of could make coordinated attacks also have kickstarted further
the aerospace industry, in the Asia Pacific region are con- R&D investment in drone neutralization technologies. The
sidered to be the major source of the growth potential. Latin ever increasing accessibility to raw explosives for building
American and African CUS markets are still in their embry- bomblets is another indirect driver of CUS market growth,
onic stages and do not show the potential for robust growth. enabling malevolent culprits to build and detonate DIY
However, the recently escalating numbers of drone attacks, bomblets. These market growth drivers are currently adding
especially in Latin America, are likely to spark governmental fuel to the explosion of the CUS market, yet not without
investments in developing CUS technologies, which would mitigating factors.
subsequently drive market growth in that region.
4) MARKET GROWTH INHIBITOR
In tandem with multiple market growth drivers, multi-
3) MARKET GROWTH DRIVER dimensional factors can exist that restrain CUS market
Aside from the obvious market needs to counteract the poten- growth (summarized in Table 8).
tial threats posed by mUAVs, the growth drivers of the CUS • Technology obsolescence: First, new UAV technologies
market are multi-faceted, including both direct and indirect are being developed at a rapid pace. Smaller UAVs that
antecedents (as summarized in Table 8). fly longer and are equipped with better aerial imaging

168694 VOLUME 8, 2020

H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

TABLE 8. Derivers and Inhibitors of CUS Market Growth.

units make it progressively difficult for CUS providers bullet-proof countermeasures due to the limited number
to detect and neutralize mUAVs. Knowledge in the area of UAV attacks. Akin to the proverbial firefighter in a
of mUAVs quickly becomes obsolete, which signifi- town where there are no fires, CUS providers in a market
cantly shortens the shelf life of CUS providers’ organi- where UAV attacks rarely exist would have a difficult
zational capabilities. Consequently, CUS providers are time identifying potential threats and designing a series
required to keep up with UAV innovations to launch of counteractions.
newly updated countermeasures.
• Security concerns: Second, regulatory bodies pro-
hibiting defense-related manufacturers from exporting 5) MARKET FRAGMENTATION
restrict the global expansion of CUS providers. For The civilian CUS market is currently highly fragmented;
instance, the US International Traffic in Arms Regu- there is no single dominant player which could exert suf-
lation prohibits certain American CUS manufacturers ficient influence to drive the entire industry towards its
from selling overseas. Recently, however, the US gov- intended direction. Instead, multiple established corporations
ernment started to clear CUS manufacturers on a case and diverse small- and medium-sized enterprises (SMEs)
by case basis to export CUS systems strictly to allied together constitute a dynamically evolving ecosystem. The
nations. For instance, Raytheon was approved by the marketplace of solution providers, particularly value-added
government to sell the Coyote Block 2 counter-drone resellers, is an example of an intrinsically highly fragmented
weapon to approved allied nations in March of 2020. market. Because solution providers are required to customize
• Incomplete regulation: Third, the rules of engagement each solution for individual clients, economies of scale that
to counteract mUAVs have not yet been fully developed. are usually attained through providing standardized and uni-
The US Department of Justice released a guideline to versally deployable services are unachievable in most cases.
counteract killer drones in August of 2016 based on the The requirements of individual customization and subsequent
2013 Presidential Policy Guideline to establish standard customer support for an extended period therefore naturally
operating procedures to counteract terror attacks [303]. turn away large corporations from entering the market. Local
Although specific rules of drone engagement seem to be SMEs typically fill this void by leveraging their relatively
in detail, there still exists some wiggle room that could low-cost structures compared to those of large corporations.
generate multiple interpretations. The hospitality television market is a typical example of
• Lack of assessment criteria: Fourth, from the perspec- this intrinsically highly fragmented marketplace for solution
tive of CUS clients, universal assessment criteria to providers: although many consumer electronics corporations,
evaluate CUS performance capabilities are non-existent. such as Phillips and Samsung, briefly considered entering the
This seemingly insignificant factor makes it difficult television solution market for hotels, retail outlets, doctors’
for potential clients to decide whether they need a offices, and cruise ships, none of them found the target market
CUS solution and should choose from among the many profitable enough. The less than satisfying profitability level
CUS companies providing vastly different technologies can be attributed to the intrinsic nature of this fragmented
and solutions. Without standardized assessment criteria, market, which requires individual customization and contin-
clients are left to compare apples and oranges. uous customer support. The civilian CUS market followed
• Insufficient cases for analysis: Lastly yet ironically, the footsteps of this intrinsically highly fragmented market
the general understanding of concrete threat profiles of until recently: while traditional defense corporations, such
mUAVs is hardly sufficient at the moment to develop as Lockheed Martin and the Thales Group were handling

VOLUME 8, 2020 168695

 H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

the military needs for CUS hardware and software, regional Grumman’s Drone Restricted Access Using Known EW
SMEs such as DroneShield, DeDrone, and Aveillant as new (DRAKE) CUS system [321], which is a radio fre-
entrants in the CUS market during the 2010s started to pro- quency negation system delivering a non-kinetic elec-
vide civilian applications. However, the civilian CUS market tronic attack, and Raytheon’s Coyote CUS system [270],
growth has shown a significant upsurge recently, becoming which consists of Ku-band multi-spectral detecting and
substantial enough to attract traditional defense corporations. high-energy laser neutralization functions, strictly target
In this highly fragmented yet organically intertwined market, military applications.
both established multi-national corporations and regionally • European: European triumvirate corporations in the
based SMEs do not necessarily compete with each other but, CUS market consist of the Thales Group in France,
rather create symbiotic relationships with one another. Each Saab AB in Sweden, and Blighter Surveillance Sys-
entity brings the complementary assets to the marketplace, tems Ltd. in the UK. In contrast to the military-centric
consequently creating overall both competing and cooper- product and service portfolio of the North American
ating relationships. The next section introduces the current triumvirate, European triumvirate serves both the mili-
players in the civilian CUS market and the notable acquisi- tary and civilian markets. Compared to the heavy weight
tions among them. champions of Thales and Saab, whose legacy products
and service portfolios heavily rely on military appli-
B. GAME OF DRONES: WHO ARE THE CURRENT MAJOR cations, the UK’s Blighter Surveillance System plays
PLAYERS IN THE CIVILIAN CUS MARKET? a relatively light-weight game specifically focused on
The civilian CUS market has been a dynamically evolv- the CUS market. Blighter’s Anti-UAV Defense System
ing ecosystem which consists of ‘‘big fish’’ in the ocean, (AUDS) solution combines Blighter’s A400 Series air
i.e., established multinational corporations positioned in security radar with its HawkEye video tracker armed
the traditional defense industry, such as Lockheed Martin, with a directional radio frequency inhibitor for signal
Thales, Raytheon, Saab, and BSS Holland, and ‘‘small fish’’ jamming in order to serve areas of demand in the civilian
in the pond, i.e., emerging startups and SMEs positioned in market, such as airports, nuclear power plants, and high-
local civilian CUS markets, such as Aveillant, DroneShield, end commercial compounds, in addition to defense,
Dedrone, Citadel Defense, and Liteye, as well as a special national border security, law enforcement, and coastline
batch of ‘‘small fish’’ in the pond, i.e., spinoff companies security. Blighter’s AUDS is the most ambidextrous,
from SMEs such as Fortem Technologies as a spinoff of full-stack CUS solution among those of the European
ImSAR LLC. Table 9 showcases a selective group of estab- triumvirate, consisting of Blighter’s hardware capability
lished corporations and small enterprises in the global CUS in radar technology and its proprietary software. Saab
market. In this section, big fish and small fish are analyzed and Thales, on the other hand, have a stronger presence
to understand the current ecosystem of the global civilian in terms of hardware capability. Moreover both corpo-
CUS market. Recent acquisitions between big fish and small rations also offer a wide range of UAV products and
fish which have blurred the boundaries between oceans and solutions, such as Saab’s Skeldar Series and Thales’s
regional ponds are also analyzed in this section. WatchkeeperX, Spyranger, and Fulmar models. As a
result, the prospect of the emerging and booming CUS
market poses a Catch-22 for both Saab and Thales by
1) BIG FISH IN THE OCEAN
placing both companies in the position of a locksmith
North American and European defense industries have been who is tasked with inventing both an unlockable lock
ruled by a group of dominant players. and passe-partout. An all-out war in the CUS market
• North America: The North American defense triumvi- by Saab and Thales would cannibalize their own UAV
rate, i.e., i) Lockheed Martin, ii) Northrop Grumman, markets. As a result, both Saab and Thales shy away
and iii) Raytheon, boasts expansive product and service from entering the CUS market in full force, rather adopt-
portfolios catering to the aerospace and defense indus- ing an alternative route of focusing on the detecting
tries. All three corporations have a strong foothold in and tracking functions of the CUS system by lever-
the military UAV industry: i) Lockheed Martin’s Indago, aging their existing radar technologies. Saab’s Giraffe
Condor, and Stalker [317], ii) Northrop Grumman’s Enhanced Low, Slow and Small (ELSS) CUS system is
Global Hawk [318], and iii) Raytheon’s Coyote and Sil- built on its Giraffe surveillance radar, whereas Thales’s
ver Fox UAVs are all-time major players in military UAV Horus Captor is built on its short-range, low-altitude
systems [270], [319]. Among the triumvirate, Lockheed surveillance radar. Thales, however, has recently went
Martin entered the civilian CUS market by introducing one step further: Thales acquired Aveillant, a UK com-
ICARUS, a Q-53 radar system that detects mUAVs and pany developing drone detection solution using holo-
triggers a kill chain to defeat targets using its Advanced graphic radar technology, in November of 2017, later
Test High Energy Asset System (ATHENA), a trans- acquiring Drone Shield, an Australian CUS solution
portable ground-based system equipped with a 30 kilo- company, in May of 2019. Mostly owing to these back-
watt laser beam [320]. In comparison, both Northrop to-back major acquisitions, Thales recently made the

168696 VOLUME 8, 2020

H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

TABLE 9. Exemplary Companies in The Global CUS Market (Alphabetic order).

biggest splash in the civilian CUS market by launching Leonard DRS, one of the top US defense contractors, with a
EagleSHIELD, a turn-key CUS solution to detect, track, $42m contract to develop CUS capability in October of 2017
and neutralize mUAVs, in November of 2019. Thales’s [322]. In 2019, the US DoD spent $900m on developing CUS
Horus Captor is integrated with Drone Shield’s neu- solutions [297]. This series of large-scale cash injections by
tralization systems, whose defeat functions range from the military sector subsequently fertilized the civilian CUS
hijacking, jamming, interception, to both electronic and market through defense contractors teaming up with CUS
physical destruction. Thales’s acquisitions reflect the firms as suppliers. For instance, Liteye Systems partnered
inherent nature of highly fragmented markets: numerous with Northrop Grumman to combine Liteye’s counter UAS
small fish in local ponds. defense system with Northrop Grumman’s Stryker Infantry
Carrier Vehicle to create a combat-level, powerful CUS solu-
2) SMALL FISH IN THE POND tion. The majority of small enterprises in the CUS mar-
The 2010s witnessed the simultaneous sprouting of new ket possess hardware technologies that offer a competitive
entrants in the CUS market; rather dormant market of CUS advantage, such as DroneShield’s proprietary acoustic detec-
products and services suddenly became crowded with US tion technology, Aveillant’s holographic radar technology,
firms, such as Aveillant, Dedrone, DroneShield, Airpsace and Drone Defense’s solar-powered off-grid radio frequency
Systems, SkySafe, Citadel Defense, WhiteFox Defense Tech- scanning and detection technology [323], to name a few.
nologies and Liteye Systems. Table 9 summarizes the CUS A subgroup of firms armed with strong hardware capabil-
product and service portfolios of these incumbents. Concur- ities specifically focus on neutralization technology, such as
rent with the active market dynamism in the civilian CUS the DroneHunter interceptor drone by Fortem Technologies
market, the entire CUS ecosystem became more vibrant with equipped with its interdependent subsystem known as Drone-
large-scale defense contracts, such as the US Army awarding Hangar, a charging deck, and a netting gun.

VOLUME 8, 2020 168697

 H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

FIGURE 9. Map of CUS market dynamics showing acquisitions (−− →) and partnerships (← −− →).

On the other hand, a group of small enterprises in the CUS Fig. 9 visualizes the prototypical pathway that SMEs take to
market proudly has a strong competitive advantage in terms of reach global markets by partnering with established corpora-
software capabilities. For instance, Dedrone’s DroneTracker tions or even acquiring certain divisions of established corpo-
software is built based on a machine learning-based algorithm rations. For instance, Dedrone, whose competitive advantage
to recognize and classify mUAVs. Dedrone also developed a is significantly lopsided given its strong software capability,
proprietary database of all UAVs currently available in both purchased all assets and intellectual properties associated
military and civilian markets. SkySafe’s Airspace Galaxy is with Batelle’s DroneDefender, a 15-lb man-portable CUS
also built on top of machine learning-based algorithms to shoulder rifle, in October of 2019 [325]. Dedrone’s product
detect and identify mUAVs. Citadel Defense’s Titan is built and service portfolio initially leaned heavily towards detect-
on top of its artificial-intelligence-based proprietary algo- ing and tracking mUAVs using Dedrone’s machine learning-
rithms. based DroneTracker software integrated with radio frequency
These asymmetric competitive advantages among small sensors and pan-tilt-zoom cameras. SMEs strike partnerships
enterprises in the CUS market are mostly due to the lim- with established corporations for a certain project or prod-
ited level of slack resources, including both tangible and uct/service portfolio to extend their global reach, as repre-
intangible assets. This asymmetry in competitive advantages sented by Pathway 3 in Fig. 9. Liteye’s series of partnerships
motivates small enterprises to search for potential partner- with established aerospace and defense corporations, such as
ships with existing or potential competitors whose resources Raytheon and Northrop Grumman, are other examples which
and capabilities would complement their own. Partnerships represent partnerships between small enterprises armed with
between small enterprises with complementary capabilities customizable solutions and established corporations possess-
are most common. For instance, Citadel Defense recently ing advanced technology-based products. Liteye partnered
complemented its weakness in hardware capability by pitch- with Northrop Grumman to integrate Liteye’s AUDS into
ing its strength in CUS software and striking a partnership Northrop Grumman’s armored vehicle, the Stryker Infantry
with Liteye, whose strength is in hardware technologies, Carrier Vehicle, and introduced the integrated system at
in March of 2020 [324]. Pathway 1 in Fig. 9 represents the the US Army’s Maneuver and Fires Integration Exercise in
prototypical pathway that SMEs with different competitive November of 2018 [326]. Most recently, Liteye teamed up
advantages take by partnering with competitors whose com- with Raytheon Missile & Defense to integrate Liteye’s AUDS
petitive advantages complement their own. In addition to with Raytheon’s PhaserTM high-powered microwave sys-
these types of partnerships among small incumbents in the tem in April of 2020 [327]. Dedrone’s acquisition of Bet-
CUS market, there also exists a particular group of potential telle’s neutralization system or Liteye’s partnerships with
competitors which could help small enterprises to expand to Raytheon and Northrop Grumman showcase partnerships
the global scale: the big fish in the ocean. between regional enterprises with a specific yet short range
of capabilities and established or even multinational cor-
3) SMALL FISH MOVING FORWARD TO THE OCEAN porations with a diverse portfolio of products and services
THROUGH BIG FISH where regional enterprises seize lucrative opportunities to
Those small enterprises that strike partnerships with estab- move forward to larger market while established corporations
lished corporations are likely to gain a foothold quickly relatively inexpensively acquire necessary and highly specific
to expand themselves into larger markets. Pathway 2 of assets. Occasionally partnerships between small enterprises

168698 VOLUME 8, 2020

H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

and established corporations are leveraged in the opposite Therefore, incorporating the unique characteristics of the
direction, where a big fish attempts to reach regional yet network type (i.e., MANET, VANET, and FANET) and the
highly lucrative ponds by swallowing smaller fish. network role, integrated CUS networks must be flexible
and manageable. These networks must balance/optimize the
4) BIG FISH REACHING THE POND THROUGH SMALL FISH requirements, e.g., platform mobility, robust transmission
Established corporations generally expand into solution mar- links, delay, scalability, multiple access schemes, and limited
kets by acquiring regional SMEs with a proven track record. resource allocation. To deal with CUS network optimization,
Pathway 4 in Fig. 9 presents the prototypical route used by software-defined networking (SDN) and network function
established corporations in the aerospace and defense indus- virtualization (NFV) technologies can effectively manage the
try target to move into to the CUS market, i.e., by acqui- resources for an inter-operable CUS with various platforms
sition. Thales acquired Aveillant, the British CUS company [333], [334]. It is worth noting that the UAVs in a CUS
with proprietary holographic radar technology, in Novem- network not only be benefitted by SDN/NFV, but also support
ber of 2017 and subsequently established Aveillant Limited other platforms as programmable network nodes [56], [335]–
Thales Company Operational Solutions Ltd. [328]. Shortly [337] (see [333] and references therein for the survey of
before the acquisition, Aveillant had proved its competency UAV related SDN/NFV). For example, SDN and NFV can
in the civilian CUS market by installing its Gamekeeper handle flexible power allocation, coordination of the band-
CUS system in Monaco on April of 2016 [329]. Aveillant’s width/channel allocation, and routing algorithms. Unified
grand entrance into the global CUS market immediately management and optimization of the dynamic configuration
sparked clients’ interest in CUS solutions. In April of 2017, of the networks can be realized through SDN and NFV. SDN
Singapore’s ST Electronics installed Aveillant’s Gamekeeper decouples the control plane from the data plane and simplifies
near the Singapore Flyer attraction [330]. Paris Charles de network management and control [338], [339]. NFV decou-
Galle Airport in Paris was the third international VIP client ples the hardware and software and enhances the flexibility
of Aveillant, installing Gamekeeper in July of 2017 [331]. of the networks by using virtualization techniques [340]. It is
Three major international installations provided Thales with noteworthy that network slicing can also achieve flexible and
sufficient validation to its decision to acquire Aveillant. manageable networks that can be realized with SDN and
Shortly after the Aveillant acquisition, Thales then acquired NFV. Network slicing refers to a logical network that can
DroneShield, an Australian CUS company specialized in provide mission-specific capabilities [341]. Satellite commu-
acoustic UAV detection technology, in May of 2019 [332]. nications beyond HAPs can also be considered to imple-
ment SDN-based high-performance communications. Het-
VII. CHALLENGES AND FUTURE DIRECTION erogeneous satellite communication networks can be interop-
In this section, we present the challenges and future direction erable with ground/sky platforms and can provide flexibility
for the CUS research and the vision of the CUS market. according to the service requirements based on SDN and
NFV [342], [343].
A. LIMIT AND CHALLENGES OF CUS NETWORKS CUS networks need to be designed carefully to satisfy mul-
A single platform can hardly cope with unpredictable threats tiple objectives with the capabilities of flexibility and man-
from emerging mUAVs, and multiple platforms in CUS net- ageability. These networks can be centralized/decentralized
works providing diversity and reliability will be a promis- and homogeneous/heterogeneous to balance the tradeoff
ing solution. Intercommunication between the platforms is between robustness and performance; however, their archi-
a critical factor that allows the system effectively to operate tecture is fixed which limits their performance. Emerg-
integrated CUS and CUS networks. Proper network and com- ing technologies such as SDN, NFV, network slicing, and
munication performance satisfying mission requirements is resource optimization will enable these networks to be flex-
the key issue to maximize the performance of an integrated ible and manageable in terms of communications and net-
CUS network. working performances.
Integrated CUS networks can consist of static/mobile
ground platforms, human-packable platforms, and high/low
altitude sky platforms. These networks among (quasi-)static
ground platforms are represented as a mobile ad hoc net- B. DEARTH OF ASSESSMENT CRITERIA FOR CUSs
work (MANET); networks among mobile ground platforms To tackle with dearth of assessment criteria which was
are represented as a vehicle ad hoc network (VANET); and explained in Section VI-A, several objectives as well as ana-
networks among sky platforms are represented as a flying lytical and experimental studies can be investigated. We pro-
ad hoc network (FANET). These respective ad hoc networks pose a few performance objectives including mUAV neutral-
have unique characteristics in terms of mobility, topology, ization probability (mUNP), expected loss of profit (ELP),
topology changes, energy constraints, and uses [69]. Further- covering space per cost (CSC), mitigation completion time
more, networks having different objectives, e.g., detection, (MCT), mitigation completion power (MCP), capacity of
computation, and neutralization, have different communica- mitigation (COM), mitigation cycle of CUS (MCC), and
tion requirements, e.g., rates, latencies, and mobility levels. operating duration of CUS (ODC), as follows:

VOLUME 8, 2020 168699

 H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

• mUNP: It is defined as Pd Pm where Pd and Pm are gies, such as wireless communications, networks, con-
detection and mitigation probabilities, respectively. The trol theory, mechanics, and computer science, more
mUNP then represents the success probability of CUS comprehensive frameworks are required to effectively
mission. An instantaneous mUNP can be maximized by integrate them.
allocating resources, e.g., power, spectrum, and sensing • Dynamic and flexible CUS networks: Since each plat-
time, to the devices and functions for the given CUS form has unique benefits and limitations in terms of the
platforms and architectures, and an average mUNP can mobility, topology, energy cases, and usages, a dynamic
be employed by designing and deploying the CUS plat- and flexible CUS network is desired to significantly
forms and architectures. When an mUAV approaches improve the CUS performance. For example, network-
from a blind spot, we can consider the worst-case mUNP ing the sensing, C2, and mitigation systems to flatten
and maximize it so that lower bound of neutralization the hierarchy, reduce the operational pause, enhance
performance is increased, resulting ina robust
 CUS. precision, and increase the response speed of command
• ELP: It is defined
  as (1 − Pd Pm )E Cp + P f E [Cc ], [156]. Here, SDN/NFV could be a relevant solution to
where E Cp denotes the expected damage cost in a establish the dynamic and flexible networks.
protective area, Pf denotes a false alarm probability, and • Fusion or confusion?: The CUS consists of various sen-
E [Cc ] denotes expected collateral damage caused by the sors, each of which collects and provides heterogeneous
false alarm. The ELP represents the expected net cost by information. Direct merge of various data with the lack
operating the CUS, which should be minimized in the of caution may confuse rather than clarify the decision
design of CUS. of CUS. The heterogeneous data should be intelligently
• CSC: It is space that can be covered by a single CUS fused to establish robust and effective sensing systems
with the normalized cost of resources. In other words, that can correctly detect/identify, authorize, localize, and
CSC represents how broad space can be effectively cov- track the mUAVs. For the intelligent fusion of sensing
ered by a CUS by using normalized resources, such data, recently and dramatically developed artificial intel-
as power, spectrum, and sensing time. In open space, ligent (AI) technologies could enhance performance of
the CSC is equivalent to the maximum range (distance) data fusion.
of the effective mitigation of CUS per unit resource use. • Automation and fast computation: MCT is a critical
• MCT/MCP: It is a required time/power to successively factor to protect skies. Inefficient computing strategies
mitigate a single mUAV. Based on MCT/MCP, we can of C2 systems as well as human interventions cause
predict the required time/power consumption to mitigate significant latency and large MCT. To reduce the com-
multiple mUAVs and to complete a mission in various puting time, edge computing can be employed in which
attack scenarios of mUAV. nearby edge nodes provide computation to CUS. Edge
• COM: It is the largest number of mUAVs that CUS can computing can also provide stronger security and better
simultaneously mitigate. COM can be used to determine interoperability. On the other hand, AI can minimize
how many CUSs are required to cover the target area and human interventions to enhance CUS performance.
how to schedule them to effectively protect the target • Catch me if you can: The rapid and innovative devel-
area. opment of UAVs make the existing CUS obsolete.
• MCC: It is a number of mUAVs that CUS can mitigate To prepare for every eventualities including attacks from
per unit time. MCC can be used with COM to design the mUAVs performing cyberattacks (e.g., GNSS/RF jam-
defense systems. ming and spoofing), the knowledge of the state-of-the-
• ODC: It is an operating duration that CUS can con- art mUAV is required. It is however challenging to iden-
tinuously operate without recharging or returning to a tify all types of mUAVs and obtain the information of the
base. These performance criteria should be studied and mUAVs. Therefore, the physical mitigation which uses
employed according to several specific scenarios (e.g., less or none of knowledge about mUAVs is required (see
24/7-operation requirement and ultra-sensitive areas). Section V.B.2). For the physical mitigation, the mUAV
tracking and chasing algorithms should be developed
C. TECHNOLOGICAL CHALLENGES AND STRATEGIES as stated in Section III.B. Also, the fast and accurate
mobility of mobile platforms (e.g., ground mobile and
To enhance the performance of CUSs and achieve the objec-
sky platforms) should be studied.
tives introduced in the previous subsection, there are still
• Price reduction: The price of UAVs is decreasing
numerous technological challenges remaining. The techno-
and more affordable, whereas the price of CUSs is
logical challenges/issues and strategies to achieve/resolve
much more expensive than UAVs. The asymmetric cost
them are summarized as follows:
between UAV and CUS would hinder the defenders
• Fundamental framework and prototype for CUS: from protecting wide exposed area. Therefore, improv-
Fundamental framework for CUS has not been fully ing the energy efficiency of CUSs is important to make
characterized and analyzed in both academia and indus- CUS sustainable and reusable to cover wide area. Fur-
tries. Since CUS is an integration of various technolo- thermore, developing the low-cost sensors/mitigators is

168700 VOLUME 8, 2020

H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

TABLE 10. Yin and Yang of the UAV and CUS Industries: What are the Lessons for Adjacent Market Players?.

critical to reduce the price of CUS, so that wide area can of the UAV industry are also expected to be substanti-
be protected with low cost. ated, which consequently would facilitate the growth of
the CUS market. On the other hand, if UAV regulations
D. YIN AND YANG OF THE CUS INDUSTRY: WHAT ARE move towards more conservative and strict domains,
THE LESSONS FOR ADJACENT MARKET PLAYERS? the negative externalities of the UAV industry would
The rise of the UAV industry has become the backdrop of automatically be contained, thus inhibiting the growth
a modern-day gold rush: multiple stakeholders ranging from of the CUS market.
manufacturers to service providers have quickly filled the • Lastly, tapping into both the UAV and CUS markets
emerging ecosystem of UAVs. One group of the stakeholders will create the ultimate Catch-22 of cannibalizing one’s
has arrived at the scene with malevolent intentions and has own product and service portfolio. These distinct char-
abused the newly developing UAV innovations. The global acteristics of the global CUS market pose unique oppor-
CUS market blossomed as a direct response to this unmet tunities for industry players in adjacent markets, such
market need that had been created by negative externalities as telecommunication service providers, consumer elec-
of the UAV industry. As a result, the CUS market shows a tronics companies, and software companies. Although
few characteristics that are distinctively different from the still highly uncertain, potential entrants to the CUS
majority of newly created technology markets (as summa- market, especially established corporations, may have
rized in Table 10). to draw an analogy from large corporations in the
• First, the current market needs in the CUS market aerospace and defense industry, such as Lockheed Mar-
entirely depend on the activities, especially those with tin, Northrop Grumman, and Thales. Due to the lim-
negative impacts, in the UAV industry. On the other ited nature of market opportunities in the CUS market,
hand, the UAV industry also partially depends on the as explained in this section, corporate entrants should
activities in the CUS market to a certain degree, although heed Thales’s strategy of acquiring regional small enter-
the level of dependence is much lower than that of the prises with strong capabilities in integrated CUS solu-
CUS market. This mutual dependence creates a type tions.
of yin and yang dynamics between the UAV industry We also investigate the new industry emergence and market
and the CUS market: one system’s activities have direct dynamics of CUS in this study. Our findings in this section
impacts on the other system, and vice versa, and one highlight the emergence process of the CUS industry and
system can hardly exist independently without the other the distinctive characteristics of the current CUS market.
system’s prosperity. However, it is quite a stretch to label Due to the limited number of market incumbents which is
the UAV and CUS market as yin and yang dynamics the innate limitation of the newly emerging industry and
owing to the mutual yet asymmetric interdependence market, however, qualitative analysis approach was the only
that exists. viable option for this study. As the CUS industry evolves
• This creates the second characteristic of the CUS mar- and more incumbents enter the CUS market, future research
ket: temporal precedence of the UAV industry, which would be able to leverage diverse methodologies including
is necessary for the CUS market to emerge. Without quantitative or mixed method. For instance, as the CUS
negative externalities in the UAV industry, the market industry matures, markets are likely to be further multi-
need for the CUS market would never have existed. layered, i.e., incumbents becoming more specialized in nar-
• Third, the market size and growth rate of the CUS market rowly focused product and service portfolios. Based on the
entirely depend on the size and growth rate of negative findings of this study, further investigation on prototypical
externalities in the UAV industry as well. Without the incumbents and their alliance networks would reveal further
perceived threats of mUAVs, the anticipated growth of market dynamism of the CUS industry.
the CUS market is highly unlikely.
• Fourth, market saturation and the demise of the CUS VIII. CONCLUSION
market in future depend not only on that of the UAV In this paper, we have provided a comprehensive survey of
industry but also on potential changes in UAV regu- CUSs based on a top-down approach. Starting with UAV
lations. If UAV regulatory changes are geared towards applications, the survey has explored the platforms, archi-
more a laissez-faireism approach, negative externalities tectures, and devices and functions of CUSs. Various types

VOLUME 8, 2020 168701

 H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

of platforms, systems, and devices have been introduced and [13] EY India. (Nov. 2019). What’s the Right Strategy to Counter Rogue
their pros and cons and associated challenging issues have Drones? [Online]. Available: https://www.ey.com/en_in/emerging-
technologies/whats-the-right-strategy-to-counter-rogue-drones
been examined. CUS platforms have been categorized as [14] P. Tokekar, J. V. Hook, D. Mulla, and V. Isler, ‘‘Sensor planning for a
ground and sky platforms with the networks connecting them symbiotic UAV and UGV system for precision agriculture,’’ IEEE Trans.
based on the operating region. A lower-level taxonomy of Robot., vol. 32, no. 6, pp. 1498–1511, Dec. 2016.
[15] H. Xiang and L. Tian, ‘‘Development of a low-cost agricultural remote
CUSs, i.e., an architecture, was then introduced, including sensing system based on an autonomous unmanned aerial vehicle
three systems, sensing, C2, and mitigation systems. In the (UAV),’’ Biosyst. Eng., vol. 108, no. 2, pp. 174–190, Feb. 2011.
lowest level of the CUS taxonomy, two essential devices for [16] P. Grippa, ‘‘Decision making in a UAV-based delivery system with impa-
tient customers,’’ in Proc. IEEE/RSJ Int. Conf. Intell. Robots Syst. (IROS),
CUSs, sensors and mitigators, were reviewed. Finally, we sur- Oct. 2016, pp. 5034–5039.
veyed the CUS market and revealed its dynamics and unique [17] K. T. San, S. J. Mun, Y. H. Choe, and Y. S. Chang, ‘‘UAV delivery
characteristics. From this survey, we have identified rapidly monitoring system,’’ in Proc. Asia Conf. Mech. Aerosp. Eng. (ACMAE),
vol. 151, Feb. 2018, p. 04011.
and dynamically growing studies and businesses related to [18] N. K. Yang, K. T. San, and Y. S. Chang, ‘‘A novel approach for real time
CUSs. We believe that this in-depth survey of CUSs provides monitoring system to manage UAV delivery,’’ in Proc. 5th IIAI Int. Congr.
a timely and unique guideline for CUS development, regula- Adv. Appl. Informat. (IIAI-AAI), Jul. 2016, pp. 1054–1057.
[19] B. D. Song, K. Park, and J. Kim, ‘‘Persistent UAV delivery logistics:
tion implementation, and industry collaboration. MILP formulation and efficient heuristic,’’ Comput. Ind. Eng., vol. 120,
pp. 418–428, Jun. 2018.
ACKNOWLEDGMENT [20] S. Bang, H. Kim, and H. Kim, ‘‘UAV-based automatic generation of high-
resolution panorama at a construction site with a focus on preprocessing
The authors wish to express their deep appreciation for the for image stitching,’’ Autom. Construct., vol. 84, pp. 70–80, Dec. 2017.
support rendered by the members of the Next-Generation [21] Z. Shang and Z. Shen, ‘‘Real-time 3D reconstruction on construction site
Unmanned Vehicle Wireless Communication Laboratory using visual SLAM and UAV,’’ Dec. 2017, arXiv:1712.07122. [Online].
Available: http://arxiv.org/abs/1712.07122
(including the Mobile Communication Laboratory and the [22] M. Alzenad, A. El-Keyi, F. Lagum, and H. Yanikomeroglu, ‘‘3-D place-
Intelligent Wireless Systems Laboratory) at Chung-Ang Uni- ment of an unmanned aerial vehicle base station (UAV-BS) for energy-
versity. In particular, the authors would also like to thank efficient maximal coverage,’’ IEEE Wireless Commun. Lett., vol. 6, no. 4,
pp. 434–437, Aug. 2017.
Saquib Khan, Jimin Lee, and Hangyeol Lee for their help [23] J. Lyu, Y. Zeng, R. Zhang, and T. J. Lim, ‘‘Placement optimization of
with the collection and summarizing of some data parts in UAV-mounted mobile base stations,’’ IEEE Commun. Lett., vol. 21, no. 3,
Sections II, V, and VI. pp. 604–607, Mar. 2017.
[24] C. T. Cicek, H. Gultekin, B. Tavli, and H. Yanikomeroglu, ‘‘UAV
base station location optimization for next generation wireless net-
REFERENCES works: Overview and future research directions,’’ in Proc. 1st Int. Conf.
[1] S. Herrick. (Nov. 2017). What’s the Difference Between a Drone, Unmanned Vehicle Syst.-Oman (UVS), Feb. 2019, pp. 1–6.
UAV, and UAS? [Online]. Available: https://botlink.com/blog/whats-the- [25] Y. Zeng, R. Zhang, and T. J. Lim, ‘‘Wireless communications with
difference-between-a-drone-uav-and-uas unmanned aerial vehicles: Opportunities and challenges,’’ IEEE Com-
[2] M. Germen, ‘‘Alternative cityscape visualisation: Drone shooting as a mun. Mag., vol. 54, no. 5, pp. 36–42, May 2016.
new dimension in urban photography,’’ in Proc. Electron. Vis. Arts, [26] S. Zhang, H. Zhang, Q. He, K. Bian, and L. Song, ‘‘Joint trajectory
Jul. 2016, pp. 150–157. and power optimization for UAV relay networks,’’ IEEE Commun. Lett.,
[3] E. Kaufmann, M. Gehrig, P. Foehn, R. Ranftl, A. Dosovitskiy, V. Koltun, vol. 22, no. 1, pp. 161–164, Jan. 2018.
and D. Scaramuzza, ‘‘Beauty and the beast: Optimal methods meet [27] Y. Chen, W. Feng, and G. Zheng, ‘‘Optimum placement of UAV as
learning for drone racing,’’ in Proc. Int. Conf. Robot. Automat. (ICRA), relays,’’ IEEE Commun. Lett., vol. 22, no. 2, pp. 248–251, Feb. 2018.
May 2019, pp. 690–696. [28] P. D. Bravo-Mosquera, L. Botero-Bolivar, D. Acevedo-Giraldo, and
[4] E. Kaufmann, A. Loquercio, R. Ranftl, A. Dosovitskiy, V. Koltun, H. D. Cerón-Muñoz, ‘‘Aerodynamic design analysis of a UAV for super-
and D. Scaramuzza, ‘‘Deep drone racing: Learning agile flight in ficial research of volcanic environments,’’ Aerosp. Sci. Technol., vol. 70,
dynamic environments,’’ in Proc. Conf. Robot. Learn. (CoRL), Oct. 2018, pp. 600–614, Nov. 2017.
pp. 133–145. [29] K. Kanistras, G. Martins, M. J. Rutherford, and K. P. Valavanis, ‘‘A survey
[5] T. Tozer, D. Grace, J. Thompson, and P. Baynham, ‘‘UAVs and HAPs- of unmanned aerial vehicles (UAVs) for traffic monitoring,’’ in Proc. Int.
potential convergence for military communications,’’ in Proc. IEE Colloq. Conf. Unmanned Aircr. Syst. (ICUAS), May 2013, pp. 221–234.
Mil. Satell. Commun., Jun. 2000, pp. 10-1–10-6. [30] T. Villa, F. Salimi, K. Morton, L. Morawska, and F. Gonzalez, ‘‘Devel-
[6] M. Quigley, M. A. Goodrich, S. Griffiths, A. Eldredge, and R. W. Beard, opment and validation of a UAV based system for air pollution measure-
‘‘Target acquisition, localization, and surveillance using a fixed-wing ments,’’ Sensors, vol. 16, no. 12, p. 2202, Dec. 2016.
mini-UAV and gimbaled camera,’’ in Proc. IEEE Int. Conf. Robot. [31] J. Moore, ‘‘UAV fire-fighting system,’’ U.S. Patent 2013 0 134 254 A1,
Automat., Apr. 2005, pp. 2600–2605. May 30, 2013.
[7] R. Schneiderman, ‘‘Unmanned drones are flying high in the mili- [32] S. Gallagher. (Sep. 2013). German Chancellor’s Drone, ‘Attack,’
tary/aerospace sector [special reports],’’ IEEE Signal Process. Mag., Shows the Threat of Weaponized UAVs. [Online]. Available:
vol. 29, no. 1, pp. 8–11, Jan. 2012. https://arstechnica.com/information-technology/2013/09/german-
[8] P. Iscold, G. A. S. Pereira, and L. A. B. Torres, ‘‘Development of a hand- chancellors-drone-attack-shows-the-threat-of-weaponized-uavs/
launched small UAV for ground reconnaissance,’’ IEEE Trans. Aerosp. [33] M. S. Schmidt and M. D. Shear. (Jan. 2015). A Drone, too Small
Electron. Syst., vol. 46, no. 1, pp. 335–348, Jan. 2010. for Radar to Detect, Rattles the White House. [Online]. Available:
[9] J. Y. C. Chen, ‘‘UAV-guided navigation for ground robot tele-operation https://www.nytimes.com/2015/01/27/us/white-house-drone.html
in a military reconnaissance environment,’’ Ergonomics, vol. 53, no. 8, [34] S. White. (Apr. 2015). Japanese Man Arrested for Landing Drone on
pp. 940–950, Jul. 2010. PM’s Office in Nuclear Protest. [Online]. Available: https://www.
[10] T. Coffey and J. A. Montgomery, ‘‘The emergence of mini UAVs for reuters.com/article/us-japan-nuclear-drone/japanese-man-
military applications,’’ Defense Horizons, vol. 22, p. 1, Dec. 2002. arrested-for-landing-drone-on-pms-office-in-nuclear-protest-
[11] A. Butt, S. I. A. Shah, and Q. Zaheer, ‘‘Weapon launch system design idUSKBN0NG04520150425
of anti-terrorist UAV,’’ in Proc. Int. Conf. Eng. Emerg. Technol. (ICEET), [35] R. Whittle. (Jul. 2015). Military Exercise Black Dart to Tackle
Feb. 2019, pp. 1–8. Nightmare Drone Scenario. [Online]. Available: https://nypost.com/
[12] Federal Aviation Administration. (Sep. 2020). [Online]. Available: 2015/07/25/military-operation-black-dart-to-tackle-nightmare-drone-
https://www.faa.gov/uas/resources/by_the_numbers scenario/

168702 VOLUME 8, 2020

H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

[36] E. Zorn. (Feb. 2015). We Must Ban Drones Before it’s too Late. [Online]. [59] W. DeBusk, ‘‘Unmanned aerial vehicle systems for disaster relief:
Available: https://www.chicagotribune.com/columns/eric-zorn/ct- Tornado alley,’’ in Proc. AIAA Infotech@Aerospace Conf., Apr. 2009,
drones-ban-chuy-garcia-rahm-emanuel-perspec-0302-jm-20150227- p. 3506.
column.html [60] A. Kim. (Feb. 2020). The Korea Herald. [Online]. Available:
[37] D. Cenciotti. (Aug. 2018). Video Shows What Appears to be the http://www.koreaherald.com/view.php?ud=20200227000901
First Known Drone Attack on a Head of State. [Online]. Available: [61] F. G. Costa, J. Ueyama, T. Braun, G. Pessin, F. S. Osorio, and P. A. Vargas,
https://www.businessinsider.com/drone-attack-on-venezuela-nicolas- ‘‘The use of unmanned aerial vehicles and wireless sensor network in
maduro-may-be-first-against-leader-2018-8 agricultural applications,’’ in Proc. IEEE Int. Geosci. Remote Sens. Symp.,
[38] A. H. Michel, ‘‘Counter-drone systems,’’ Center Study Drone Bard Jul. 2012, pp. 5045–5048.
College, Bard College, NY, USA, Tech. Rep., Dec. 2019. [Online]. [62] C. Ju and H. Son, ‘‘Multiple UAV systems for agricultural applications:
Available: https://dronecenter.bard.edu/files/2019/12/CSD-CUAS-2nd- Control, implementation, and evaluation,’’ Electronics, vol. 7, no. 9,
Edition-Web.pdf p. 162, Aug. 2018.
[39] Dedrone. (2020). [Online]. Available: https://www.dedrone.com [63] J. W. Rosenarchive. (Jun. 2017). Zipline’s ambitious medical drone
[40] H. Nakamura and Y. Kajikawa, ‘‘Regulation and innovation: How should delivery in Africa. MIT Technology Review. [Online]. Available:
small unmanned aerial vehicles be regulated?’’ Technol. Forecasting https://www.technologyreview.com/2017/06/08/151339/blood-from-the-
Social Change, vol. 128, pp. 262–274, Mar. 2018. sky-ziplines-ambitious-medical-drone-delivery-in-africa/
[41] C. Stöcker, R. Bennett, F. Nex, M. Gerke, and J. Zevenbergen, ‘‘Review of [64] R. L. Hotz, ‘‘In Rwanda, drones deliver medical supplies to
the current state of UAV regulations,’’ Remote Sens., vol. 9, no. 5, p. 459, remote areas,’’ Wall Street J., Dec. 2017. [Online]. Available:
May 2017. https://www.wsj.com/articles/in-rwanda-drones-deliver-medical-
[42] T. Jones, ‘‘International commercial drone regulation and drone delivery supplies-to-remote-areas-1512124200
services,’’ RAND, Santa Monica, CA, USA, Tech. Rep. RR 1718/3 RC, [65] RIC. (Jun. 2016). Drones Going Postal—A Summary of Postal Ser-
2017. vice Delivery Drone Trials. [Online]. Available: http://unmannedcargo.
[43] A. Fotouhi, H. Qiang, M. Ding, M. Hassan, L. G. Giordano, org/drones-going-postal-summary-postal-service-delivery-drone-trials
A. Garcia-Rodriguez, and J. Yuan, ‘‘Survey on UAV cellular commu- [66] MultiGP. Accessed: Sep. 12, 2020. [Online]. Available:
nications: Practical aspects, standardization advancements, regulation, https://www.multigp.com
and security challenges,’’ IEEE Commun. Surveys Tuts., vol. 21, no. 4, [67] S. Karapantazis and F. Pavlidou, ‘‘Broadband communications via high-
pp. 3417–3442, 4th Quart., 2019. altitude platforms: A survey,’’ IEEE Commun. Surveys Tuts., vol. 7, no. 1,
[44] (2020). UAV Coach. [Online]. Available: https://uavcoach.com/drone- pp. 2–31, 1st Quart., 2005.
laws/ [68] Y. Chen, H. Zhang, and M. Xu, ‘‘The coverage problem in UAV network:
[45] G. C. Birch and B. L. Woo, ‘‘Counter unmanned aerial systems testing: A survey,’’ in Proc. 5th Int. Conf. Comput., Commun. Netw. Technol.
Evaluation of VIS SWIR MWIR and LWIR passive imagers,’’ SNL-NM, (ICCCNT), Jul. 2014, pp. 1–5.
Albuquerque, NM, USA, Tech. Rep. SAND2017-0921 650791, 2017. [69] L. Gupta, R. Jain, and G. Vaszkun, ‘‘Survey of important issues in UAV
[46] B. Nassi, A. Shabtai, R. Masuoka, and Y. Elovici, ‘‘SoK-security and communication networks,’’ IEEE Commun. Surveys Tuts., vol. 18, no. 2,
privacy in the age of drones: Threats, challenges, solution mecha- pp. 1123–1152, 2nd Quart., 2016.
nisms, and scientific gaps,’’ 2019, arXiv:1903.05155. [Online]. Available: [70] N. H. Motlagh, T. Taleb, and O. Arouk, ‘‘Low-altitude unmanned aerial
http://arxiv.org/abs/1903.05155 vehicles-based Internet of Things services: Comprehensive survey and
[47] S. Samaras, E. Diamantidou, D. Ataloglou, N. Sakellariou, A. Vafeiadis, future perspectives,’’ IEEE Internet Things J., vol. 3, no. 6, pp. 899–922,
V. Magoulianitis, A. Lalas, A. Dimou, D. Zarpalas, K. Votis, P. Daras, Dec. 2016.
D. Tzovaras, ‘‘Deep learning on multi sensor data for counter [71] S. Hayat, E. Yanmaz, and R. Muzaffar, ‘‘Survey on unmanned aerial
UAV applications—A systematic review,’’ Sensors, vol. 19, no. 22, vehicle networks for civil applications: A communications viewpoint,’’
pp. 4837–4872, Nov. 2019. IEEE Commun. Surveys Tuts., vol. 18, no. 4, pp. 2624–2661, 4th Quart.,
[48] X. Shi, C. Yang, W. Xie, C. Liang, Z. Shi, and J. Chen, ‘‘Anti-drone 2016.
system with multiple surveillance technologies: Architecture, implemen- [72] N. H. Motlagh, M. Bagaa, and T. Taleb, ‘‘UAV-based IoT platform:
tation, and challenges,’’ IEEE Commun. Mag., vol. 56, no. 4, pp. 68–74, A crowd surveillance use case,’’ IEEE Commun. Mag., vol. 55, no. 2,
Apr. 2018. pp. 128–134, Feb. 2017.
[49] I. Guvenc, F. Koohifar, S. Singh, M. L. Sichitiu, and D. Matolak, ‘‘Detec- [73] O. S. Oubbati, A. Lakas, F. Zhou, M. Güneş, and M. B. Yagoubi,
tion, tracking, and interdiction for amateur drones,’’ IEEE Commun. ‘‘A survey on position-based routing protocols for flying ad
Mag., vol. 56, no. 4, pp. 75–81, Apr. 2018. hoc networks (FANETs),’’ Veh. Commun., vol. 10, pp. 29–56,
[50] G. Ding, Q. Wu, L. Zhang, Y. Lin, T. A. Tsiftsis, and Y.-D. Yao, ‘‘An ama- Oct. 2017.
teur drone surveillance system based on the cognitive Internet of Things,’’ [74] X. Cao, P. Yang, M. Alzenad, X. Xi, D. Wu, and H. Yanikomeroglu, ‘‘Air-
IEEE Commun. Mag., vol. 56, no. 1, pp. 29–35, Jan. 2018. borne communication networks: A survey,’’ IEEE J. Sel. Areas Commun.,
[51] R. Altawy and A. M. Youssef, ‘‘Security, privacy, and safety aspects of vol. 36, no. 9, pp. 1907–1926, Sep. 2018.
civilian drones: A survey,’’ ACM Trans. Cyber Phys. Syst., vol. 1, no. 2, [75] A. A. Khuwaja, Y. Chen, N. Zhao, M.-S. Alouini, and P. Dobbins, ‘‘A sur-
p. 1–25, Nov. 2016. vey of channel modeling for UAV communications,’’ IEEE Commun.
[52] R. L. Sturdivant and E. K. P. Chong, ‘‘Systems engineering baseline Surveys Tuts., vol. 20, no. 4, pp. 2804–2821, 4th Quart., 2018.
concept of a multispectral drone detection solution for airports,’’ IEEE [76] W. Khawaja, I. Guvenc, D. W. Matolak, U.-C. Fiebig, and
Access, vol. 5, pp. 7123–7138, 2017. N. Schneckenburger, ‘‘A survey of air-to-ground propagation channel
[53] C. Kennedy and J. I. Rogers, ‘‘The emergence of mini UAVs for military modeling for unmanned aerial vehicles,’’ IEEE Commun. Surveys Tuts.,
applications,’’ Int. J. Hum. Rights, vol. 19, pp. 212–227, Feb. 2015. vol. 21, no. 3, pp. 2361–2391, 3rd Quart., 2019.
[54] D. Orfanus, E. P. de Freitas, and F. Eliassen, ‘‘Self-organization as a [77] Y. Zeng, Q. Wu, and R. Zhang, ‘‘Accessing from the sky: A tutorial on
supporting paradigm for military UAV relay networks,’’ IEEE Commun. UAV communications for 5G and beyond,’’ Proc. IEEE, vol. 107, no. 12,
Lett., vol. 20, no. 4, pp. 804–807, Apr. 2016. pp. 2327–2375, Dec. 2019.
[55] G. S. L. K. Chand, M. Lee, and S. Y. Shin, ‘‘Drone based wireless mesh [78] H. Kang, J. Joung, J. Ahn, and J. Kang, ‘‘Secrecy-aware altitude opti-
network for disaster/military environment,’’ J. Comput. Commun., vol. 6, mization for quasi-static UAV base station without eavesdropper loca-
no. 4, p. 44, 2018. tion information,’’ IEEE Commun. Lett., vol. 23, no. 5, pp. 851–854,
[56] F. Xiong, A. Li, H. Wang, and L. Tang, ‘‘An SDN-MQTT based com- May 2019.
munication system for battlefield UAV swarms,’’ IEEE Commun. Mag., [79] J.-B. Seo, S. Pack, and H. Jin, ‘‘Uplink NOMA random access for UAV-
vol. 57, no. 8, pp. 41–47, Aug. 2019. assisted communications,’’ IEEE Trans. Veh. Technol., vol. 68, no. 8,
[57] K. Daniel, S. Rohde, and C. Wietfeld, ‘‘Leveraging public wireless com- pp. 8289–8293, Aug. 2019.
munication infrastructures for UAV based sensor networks,’’ in Proc. [80] M. Zolanvari, R. Jain, and T. Salman, ‘‘Potential data link candidates for
IEEE Int. Conf. Tech. Homeland Secur. (HST), Waltham, MA, USA, civilian unmanned aircraft systems: A survey,’’ IEEE Commun. Surveys
Nov. 2010, pp. 179–184. Tuts., vol. 22, no. 1, pp. 292–319, 1st Quart., 2020.
[58] H. Menouar, I. Guvenc, K. Akkaya, A. S. Uluagac, A. Kadri, and [81] O. S. Oubbati, A. Lakas, P. Lorenz, M. Atiquzzaman, and A. Jamalipour,
A. Tuncer, ‘‘UAV-enabled intelligent transportation systems for the smart ‘‘Leveraging communicating UAVs for emergency vehicle guidance
city: Applications and challenges,’’ IEEE Commun. Mag., vol. 55, no. 3, in urban areas,’’ IEEE Trans. Emerg. Topics Comput., early access,
pp. 22–28, Mar. 2017. Jul. 25, 2019, doi: 10.1109/TETC.2019.2930124.

VOLUME 8, 2020 168703

 H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

[82] H. Shakhatreh, A. H. Sawalmeh, A. Al-Fuqaha, Z. Dou, E. Almaita, [104] (2020). Rohde & Schwarz/ESG/Diehl: Guardion. [Online]. Available:
I. Khalil, N. S. Othman, A. Khreishah, and M. Guizani, ‘‘Unmanned https://guardion.eu/
aerial vehicles (UAVs): A survey on civil applications and key research [105] (2020). Thales: Gecko-M. [Online]. Available: https://www.
challenges,’’ IEEE Access, vol. 7, pp. 48572–48634, 2019. thalesgroup.com/en/gecko-m
[83] B. Alzahrani, O. S. Oubbati, A. Barnawi, M. Atiquzzaman, and [106] (2020). Aselsan: GERGEDAN. [Online]. Available:
D. Alghazzawi, ‘‘UAV assistance paradigm: State-of-the-art in applica- https://www.aselsan.com.tr/en/capabilities/electronic-warfare-
tions and challenges,’’ J. Netw. Comput. Appl., vol. 166, Sep. 2020, systems/electronik-support-and-electronic-attack-systems/gergedan-
Art. no. 102706. portable-rcied-jammer-system-vehicle-type
[84] 5G Enhancement for UAVs, document TS 22.125, 3GPP, Nov. 2019. [107] (2020). Orion Anti-Drone System. [Online]. Available: http://www.
[85] Amazon. (Aug. 2020). First Prime Air Delivery. [Online]. Available: trd.sg/product.php
https://www.amazon.com/b?node=8037720011 [108] (2020). INT-AU002 Anti UAV System. [Online]. Available:
[86] J. Joung, ‘‘Random space-time line code with proportional fairness https://4intelligence.com/product/int-au002-anti-uav-system/
scheduling,’’ IEEE Access, vol. 8, pp. 35253–35262, 2020. [109] (2020). Take Down Your Threats Tactical and Automatic Disruptors.
[87] H.-T. Ye, X. Kang, J. Joung, and Y.-C. Liang, ‘‘Joint uplink and downlink [Online]. Available: https://www.blacksagetech.com/disruptors
3D optimization of an UAV swarm for wireless-powered NB-IoT,’’ in [110] (2020). Wingman 100 Personal Drone Alarm for Police and Security Offi-
Proc. IEEE Global Commun. Conf. (GLOBECOM), Waikoloa, HI, USA, cers. [Online]. Available: https://mydefence.dk/civil-customers/mobile-
Dec. 2019, pp. 1–6. stationary-installations/wingman-100/
[88] H.-T. Ye, X. Kang, J. Joung, and Y.-C. Liang, ‘‘Optimal time allocation [111] (2020). Drone Detection & Defense Systems Dronewatcherrf. [Online].
for full-duplex wireless-powered IoT networks with unmanned aerial Available: https://detect-inc.com/drone-detection-defense-systems/
vehicle,’’ in Proc. IEEE Int. Conf. Commun. (ICC), Shanghai, China, [112] K. Osborn. (Oct. 2017). Army Fast-Tracks New Man-Packable Counter-
May 2019, pp. 1–6. Drone EW Weapons. [Online]. Available: https://defensesystems.
[89] H.-T. Ye, X. Kang, J. Joung, and Y.-C. Liang, ‘‘Optimization for com/articles/2017/10/cw/caci-ew-army-beam.aspx
full-duplex rotary-wing UAV-enabled wireless-powered IoT networks,’’ [113] (2020). Scorpion 2 Handheld Counter-Drone Technology. [Online].
IEEE Trans. Wireless Commun., vol. 19, no. 7, pp. 5057–5072, Available: https://www.whitefoxdefense.com/scorpion
Jul. 2020. [114] (2020). Telescope Portable Drone Detection Device. [Online]. Available:
[90] C. Secchi, A. Franchi, H. H. Bülthoff, and P. R. Giordano, ‘‘Bilateral tele- https://dronetrackingtechnologies.com/products.html
operation of a group of UAVs with communication delays and switching [115] R. Isaacs, Differential Games. New York, NY, USA: Wiley, 1965.
topology,’’ in Proc. IEEE Int. Conf. Robot. Automat., Saint Paul, MN, [116] Y. Ho, A. Bryson, and S. Baron, ‘‘Differential games and optimal pursuit-
USA, May 2012, pp. 4307–4314. evasion strategies,’’ IEEE Trans. Autom. Control, vol. AC-10, no. 4,
[91] D.-T. Ho, E. I. Grøtli, P. B. Sujit, T. A. Johansen, and J. B. Sousa, pp. 385–389, Oct. 1965.
‘‘Cluster-based communication topology selection and UAV path plan- [117] J. Z. Ben-Asher, S. Levinson, J. Shinar, and H. Weiss, ‘‘Trajectory shap-
ning in wireless sensor networks,’’ in Proc. Int. Conf. Unmanned Aircr. ing in linear-quadratic pursuit-evasion games,’’ J. Guid., Control, Dyn.,
Syst. (ICUAS), Atlanta, GA, USA, May 2013, pp. 59–68. vol. 27, no. 6, pp. 1102–1105, Nov. 2004.
[92] M. Mozaffari, W. Saad, M. Bennis, Y.-H. Nam, and M. Debbah, ‘‘A tuto- [118] S.-Y. Liu, Z. Zhou, C. Tomlin, and K. Hedrick, ‘‘Evasion as a team against
rial on UAVs for wireless networks: Applications, challenges, and open a faster pursuer,’’ in Proc. Amer. Control Conf., Washington, DC, USA,
problems,’’ IEEE Commun. Surveys Tuts., vol. 21, no. 3, pp. 2334–2360, Jun. 2013, pp. 5368–5373.
3rd Quart., 2019. [119] K. Margellos and J. Lygeros, ‘‘Hamilton–Jacobi formulation for reach–
avoid differential games,’’ IEEE Trans. Autom. Control, vol. 56, no. 8,
[93] A. Sanjab, W. Saad, and T. Başar, ‘‘A game of drones: Cyber-physical
pp. 1849–1861, Aug. 2011.
security of time-critical UAV applications with cumulative prospect
theory perceptions and valuations,’’ 2019, arXiv:1902.03506. [Online]. [120] J. F. Fisac, M. Chen, C. J. Tomlin, and S. S. Sastry, ‘‘Reach-avoid
Available: http://arxiv.org/abs/1902.03506 problems with time-varying dynamics, targets and constraints,’’ in Proc.
18th Int. Conf. Hybrid Syst. Comput. Control (HSCC), Seattle, WA, USA,
[94] (2020). Blighter Surveillance Systems/Chess Dynamics/Enterprise Con-
Apr. 2015, pp. 11–20.
trol Systems: Anti-UAV Defence System (AUDS). [Online]. Avail-
[121] J. Lorenzetti, M. Chen, B. Landry, and M. Pavone, ‘‘Reach-avoid
able: https://www.homelandsecurity-technology.com/projects/anti-uav-
games via mixed-integer second-order cone programming,’’ in Proc.
defence-system-auds/
IEEE Conf. Decis. Control (CDC), Miami Beach, FL, USA, Dec. 2018,
[95] (2020). Communication & Systemes/HGH/Spectracom: Boreades.
pp. 4409–4416.
[Online]. Available: https://c-s-inc.us/products/boreades/
[122] D. W. Oyler, P. T. Kabamba, and A. R. Girard, ‘‘Pursuit–evasion games
[96] (2020). MyDefence Communication: KNOX. [Online]. Available: in the presence of obstacles,’’ Automatica, vol. 65, pp. 1–11, Mar. 2016.
https://mydefence.dk/anti-drone-solutions/knox-anti-drone-solution/
[123] X. Fang, C. Wang, L. Xie, and J. Chen, ‘‘Cooperative pursuit
[97] (2015). Airbus Defence and Space: Counter UAV System. with multi-pursuer and one faster free-moving evader,’’ Jan. 2020,
[Online]. Available: https://www.airbus.com/newsroom/press- arXiv:2001.04731. [Online]. Available: http://arxiv.org/abs/2001.04731
releases/en/2015/09/counter-uav-system-from-airbus-defence-and- [124] Z. E. Fuchs, P. P. Khargonekar, and J. Evers, ‘‘Cooperative defense within
space-protects-large-installations-and-events-from-illicit-intrusion.html a single-pursuer, two-evader pursuit evasion differential game,’’ in Proc.
[98] (2019). Raytheon: Windshear. [Online]. Available: https://www. 49th IEEE Conf. Decis. Control (CDC), Atlanta, GA, USA, Dec. 2010,
raytheon.com/news/feature/bad-drone pp. 3091–3097.
[99] (2020). Prime Consulting & Technologies: Mini/Small-Range Counter- [125] W. Scott and N. E. Leonard, ‘‘Pursuit, herding and evasion: A three-agent
UAV System. [Online]. Available: https://dronemajor.net/brands/prime- model of caribou predation,’’ in Proc. Amer. Control Conf., Washington,
consulting-technologies/products/counter-uav-system DC, USA, Jun. 2013, pp. 2978–2983.
[100] (2020). Aselsan: IHTAR. [Online]. Available: https://www.aselsan. [126] E. Garcia, D. W. Casbeer, and M. Pachter, ‘‘Design and analysis of state-
com.tr/en/capabilities/air-and-missile-defense-systems/air-and-missile- feedback optimal strategies for the differential game of active defense,’’
defense-systems/ihtar-antidrone-system IEEE Trans. Autom. Control, vol. 64, no. 2, pp. 553–568, Feb. 2019.
[101] M. Laurenzis, S. Hengy, M. Hammer, A. Hommes, W. Johannes, [127] L. Liang, F. Deng, Z. Peng, X. Li, and W. Zha, ‘‘A differential game for
F. Giovanneschi, O. Rassy, E. Bacher, S. Schertzer, and J.-M. Poyet, cooperative target defense,’’ Automatica, vol. 102, pp. 58–71, Apr. 2019.
‘‘An adaptive sensing approach for the detection of small UAV: First [128] A. Pierson, Z. Wang, and M. Schwager, ‘‘Intercepting rogue robots:
investigation of static sensor network and moving sensor platform,’’ Proc. An algorithm for capturing multiple evaders with multiple pursuers,’’
SPIE, vol. 10646, pp. 197–205, Apr. 2018. IEEE Robot. Autom. Lett., vol. 2, no. 2, pp. 530–537, Apr. 2017.
[102] (2020). Applied Technology Associates: Low-Cost Counter-Unmanned [129] D. Li and J. B. Cruz, ‘‘Defending an asset: A linear quadratic
Aerial System for Targeting (LOCUST). [Online]. Available: game approach,’’ IEEE Trans. Aerosp. Electron. Syst., vol. 47, no. 2,
http://www.atacorp.com/locust.html pp. 1026–1044, Apr. 2011.
[103] (2020). Boeing: GBAD DE OTM Ground-Based Air Defense Directed [130] R. Opromolla, G. Fasano, and D. Accardo, ‘‘A vision-based approach
Energy on-the-Move. [Online]. Available: https://www.globalsecurity. to UAV detection and tracking in cooperative applications,’’ Sensors,
org/military/systems/ground/gbad-de-otm.htm vol. 18, no. 10, p. 3391, Oct. 2018.

168704 VOLUME 8, 2020

H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

[131] M. Hua, Y. Wang, Q. Wu, H. Dai, Y. Huang, and L. Yang, ‘‘Energy- [154] (2018). Sierra Nevada Corporation: Advanced Electronic
efficient cooperative secure transmission in multi-UAV-enabled wireless Warfare System—Modular (AEWS-M). [Online]. Available:
networks,’’ IEEE Trans. Veh. Technol., vol. 68, no. 8, pp. 7761–7775, https://www.sncorp.com/media/2403/snc_aews-m-product-
Aug. 2019. sheet_2018.pdf
[132] Y. Roh, S. Jung, and J. Kang, ‘‘Cooperative UAV jammer for enhancing [155] (2020). TRD Consultancy: Fixed Site Area Protection. [Online]. Avail-
physical layer security: Robust design for jamming power and trajectory,’’ able: http://www.trd.sg/images/fixed_site_solution_brochure.pdf
in Proc. IEEE Mil. Commun. Conf. (MILCOM), Norfolk, VA, USA, [156] A. Dekker, ‘‘A taxonomy of network centric warfare architectures,’’ in
Nov. 2019, pp. 464–469. Proc. Syst. Eng./Test Eval. (SETE), Brisbane, QLD, Australia, 2008,
[133] K. Li, R. C. Voicu, S. S. Kanhere, W. Ni, and E. Tovar, ‘‘Energy efficient pp. 1–15.
[157] N. Gageik, P. Benz, and S. Montenegro, ‘‘Obstacle detection and colli-
legitimate wireless surveillance of UAV communications,’’ IEEE Trans.
sion avoidance for a UAV with complementary low-cost sensors,’’ IEEE
Veh. Technol., vol. 68, no. 3, pp. 2283–2293, Mar. 2019.
Access, vol. 3, pp. 599–609, 2015.
[134] S. Jeong, O. Simeone, and J. Kang, ‘‘Mobile edge computing via a [158] J. S. G. Guerrero, A. F. C. González, J. I. H. Vega, and L. A. N. Tovar,
UAV-mounted cloudlet: Optimization of bit allocation and path plan- ‘‘Instrumentation of an array of ultrasonic sensors and data processing
ning,’’ IEEE Trans. Veh. Technol., vol. 67, no. 3, pp. 2049–2063, for unmanned aerial vehicle (UAV) for teaching the application of the
Mar. 2018. Kalman filter,’’ Procedia Comput. Sci., vol. 75, pp. 375–380, Jan. 2015.
[135] H. Ahmadi, K. Katzis, and M. Z. Shakir, ‘‘A novel airborne self- [159] D. G. Davies, R. C. Bolam, Y. Vagapov, and P. Excell, ‘‘Ultrasonic sensor
organising architecture for 5G+ networks,’’ in Proc. IEEE 86th for UAV flight navigation,’’ in Proc. 25th Int. Workshop Electr. Drives,
Veh. Technol. Conf. (VTC-Fall), Toronto, ON, Canada, Sep. 2017, Optim. Control Electr. Drives (IWED), Moscow, Russia, Jan. 2018,
pp. 1–5. pp. 1–7.
[136] Y. Zeng, J. Xu, and R. Zhang, ‘‘Energy minimization for wireless com- [160] M. F. B. Misnan, N. M. Arshad, and N. A. Razak, ‘‘Construction sonar
munication with rotary-wing UAV,’’ IEEE Trans. Wireless Commun., sensor model of low altitude field mapping sensors for application on
vol. 18, no. 4, pp. 2329–2345, Apr. 2019. a UAV,’’ in Proc. IEEE 8th Int. Colloq. Signal Process. Appl., Melaka,
[137] Y. Zeng and R. Zhang, ‘‘Energy-efficient UAV communication with Malaysia, Mar. 2012, pp. 446–450.
trajectory optimization,’’ IEEE Trans. Wireless Commun., vol. 16, no. 6, [161] A. Al-Hourani, S. Kandeepan, and A. Jamalipour, ‘‘Modeling air-to-
pp. 3747–3760, Jun. 2017. ground path loss for low altitude platforms in urban environments,’’
[138] B. Galkin, J. Kibilda, and L. A. DaSilva, ‘‘UAVs as mobile infrastruc- in Proc. IEEE Global Commun. Conf., Austin, TX, USA, Dec. 2014,
ture: Addressing battery lifetime,’’ IEEE Commun. Mag., vol. 57, no. 6, pp. 2898–2904.
pp. 132–137, Jun. 2019. [162] H. F. Durrant-Whyte, ‘‘Sensor models and multisensor integration,’’
[139] O. M. Bushnaq, M. A. Kishk, A. Çelik, M.-S. Alouini, and in Autonomous Robot Vehicles. New York, NY, USA: Springer, 1990,
T. Y. Al-Naffouri, ‘‘Optimal deployment of tethered drones for maximum pp. 73–89.
cellular coverage in user clusters,’’ 2020, arXiv:2003.00713. [Online]. [163] B. V. Dasarathy, ‘‘Sensor fusion potential exploitation-innovative archi-
Available: http://arxiv.org/abs/2003.00713 tectures and illustrative applications,’’ Proc. IEEE, vol. 85, no. 1,
[140] (2019). Kratos Defense & Rocket Support Services: Aethon. pp. 24–38, Jan. 1997.
[Online]. Available: https://www.kratosdefense.com/systems-and- [164] E. Blasch and D. A. Lambert, High-Level Information Fusion Manage-
platforms/unmanned-systems/aerial/aethon ment and Systems Design, 1st ed. Norwood, MA, USA: Artech House,
[141] AerialX. (Apr. 2020). Dronebullet. [Online]. Available: 2012.
https://www.aerialx.com/ [165] M. Liggins, II, D. Hall, and J. Llinas, Handbook of Multisensor Data
[142] J. Mannes. (Nov. 2016). Airspace Systems: Interceptor Can Fusion: Theory and Practice, 2nd ed. Boca Raton, FL, USA: CRC press,
Catch High-Speed Drones All by Itself. [Online]. Available: 2008.
https://techcrunch.com/2016/11/18/airspace-systems-interceptor-can- [166] F. Castanedo, ‘‘A review of data fusion techniques,’’ Sci. World J.,
catch-high-speed-drones-all-by-itself/ vol. 2013, pp. 1–19, Oct. 2013.
[143] Z. Alkhalisi. (Nov. 2016). Exponent: Drone Hunter. [Online]. Available: [167] X. Shi, C. Yang, W. Xie, C. Liang, Z. Shi, and J. Chen, ‘‘Anti-drone
https://money.cnn.com/2016/11/04/technology/dubai-airport-drone- system with multiple surveillance technologies: Architecture, implemen-
hunter/ tation, and challenges,’’ IEEE Commun. Mag., vol. 56, no. 4, pp. 68–74,
[144] J. Plaza. (Jun. 2017). ALX Systems Announces New Operating Apr. 2018.
System at Commercial UAV Expo Europe. [Online]. Available: [168] S. Park, S. Shin, Y. Kim, E. T. Matson, K. Lee, J. C. Slater, M. Scherreik,
https://www.commercialuavnews.com/infrastructure/alx-systems- M. Sam, J. C. Gallagher, P. J. Kolodzy, B. R. Fox, and M. Hopmeier,
new-operating-system-uav-europe ‘‘Combination of radar and audio sensors for identification of rotor-
[145] K. D. Atherton. (Dec. 2018). Russia’s Carnivora is Designed type unmanned aerial vehicles (UAVs),’’ in Proc. IEEE Sensors, Busan,
for a Drone-Eat-Drone World. [Online]. Available: https://www. South Korea, Nov. 2015, pp. 1–4.
c4isrnet.com/unmanned/2018/12/14/russias-carnivora-is-designed-for-a- [169] G. L. Charvat, A. J. Fenn, and B. T. Perry, ‘‘The MIT IAP radar course:
drone-eat-drone-world/ Build a small radar system capable of sensing range, Doppler, and syn-
[146] (1990). Lockheed Martin: Blackbird. [Online]. Avail- thetic aperture (SAR) imaging,’’ in Proc. IEEE Radar Conf., Atlanta, GA,
able: https://www.lockheedmartin.com/en-us/news/features/ USA, May 2012, pp. 138–144.
history/blackbird.html [170] E. Diamantidou, A. Lalas, K. Votis, and D. Tzovaras, ‘‘Multimodal deep
[147] (2015). General Atomics: MQ-9 Reaper. [Online]. Available: learning framework for enhanced accuracy of UAV detection,’’ in Proc.
https://www.af.mil/About-Us/Fact-Sheets/Display/Article/104470/mq- Int. Conf. Comput. Vis. Syst. Thessaloniki, Greece: Springer, Sep. 2019,
9-reaper/ pp. 768–777.
[148] (2015). General Atomics: MQ-1B Predator. [Online]. Available: [171] F. Fioranelli, M. Ritchie, H. Borrion, and H. Griffiths, ‘‘Classification of
https://www.af.mil/About-Us/Fact-Sheets/Display/Article/104469/mq- loaded/unloaded micro-drones using multistatic radar,’’ Electron. Lett.,
1b-predator/ vol. 51, no. 22, pp. 1813–1815, Oct. 2015.
[149] (2014). Northrop Grumman: RQ-4 Global Hawk. [Online]. Available: [172] M. Ritchie, F. Fioranelli, H. Borrion, and H. Griffiths, ‘‘Multistatic micro-
https://www.af.mil/About-Us/Fact-Sheets/Display/Article/104516/rq-4- Doppler radar feature extraction for classification of unloaded/loaded
global-hawk/ micro-drones,’’ IET Radar, Sonar Navigat., vol. 11, no. 1, pp. 116–124,
[150] (2020). General Atomics: Gray Eagle UAS. [Online]. Available: Jan. 2017.
http://www.ga-asi.com/gray-eagle [173] Y. Gu, A. Lo, and I. Niemegeers, ‘‘A survey of indoor positioning systems
[151] Google X: Project Loon. Accessed: Sep. 12, 2020. [Online]. Available: for wireless personal networks,’’ IEEE Commun. Surveys Tuts., vol. 11,
https://x.company/projects/loon/ no. 1, pp. 13–32, 1st Quart., 2009.
[152] (2015). ECA Group/Group Gorge: IT180 Drone. [Online]. Available: [174] F. Zafari, A. Gkelias, and K. K. Leung, ‘‘A survey of indoor localization
https://www.ecagroup.com/en/event/neutralization-malicious-drones- systems and technologies,’’ IEEE Commun. Surveys Tuts., vol. 21, no. 3,
eca-group-innovating-and-validates-unique-technology-locate pp. 2568–2599, 3rd Quart., 2019.
[153] M. Tewksbury. (Jun. 2019). Raytheon: Howler. [Online]. Available: [175] J. Joung, S. Jung, S. Chung, and E.-R. Jeong, ‘‘CNN-based Tx–Rx dis-
http://raytheon.mediaroom.com/2019-06-18-US-Army-deploys-Howler- tance estimation for UWB system localisation,’’ Electron. Lett., vol. 55,
counter-UAS-capability-into-the-battlefield no. 17, pp. 938–940, Aug. 2019.

VOLUME 8, 2020 168705

 H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

[176] E. Mazidi, ‘‘Introducing new localization and positioning system for [201] V. Phipatanasuphorn and P. Ramanathan, ‘‘Vulnerability of sensor net-
aerial vehicles,’’ IEEE Embedded Syst. Lett., vol. 5, no. 4, pp. 57–60, works to unauthorized traversal and monitoring,’’ IEEE Trans. Comput.,
Dec. 2013. vol. 53, no. 3, pp. 364–369, Mar. 2004.
[177] I. A. Mantilla-Gaviria, G. Galati, M. Leonardi, and [202] P. Nguyen, M. Ravindranatha, A. Nguyen, R. Han, and T. Vu, ‘‘Investigat-
J. V. Balbastre-Tejedor, ‘‘Time-difference-of-arrival regularised location ing cost-effective RF-based detection of drones,’’ in Proc. 2nd Workshop
estimator for multilateration systems,’’ IET Radar, Sonar Navigat., vol. 8, Micro Aerial Vehicle Netw., Syst., Appl. Civilian Use (DroNet), Singapore,
no. 5, pp. 479–489, Jun. 2014. Jun. 2016, pp. 17–22.
[178] Y. Wang and K. C. Ho, ‘‘TDOA positioning irrespective of source range,’’ [203] P. Nguyen, H. Truong, M. Ravindranathan, A. Nguyen, R. Han, and
IEEE Trans. Signal Process., vol. 65, no. 6, pp. 1447–1460, Mar. 2017. T. Vu, ‘‘Matthan: Drone presence detection by identifying physical sig-
[179] W. Dargie and C. Poellabauer, Fundamentals of Wireless Sensor Net- natures in the drone’s RF communication,’’ in Proc. 15th Annu. Int.
works: Theory and Practice. Hoboken, NJ, USA: Wiley, 2010. Conf. Mobile Syst., Appl., Services, Niagara Falls, NY, USA, Jun. 2017,
[180] S. Basak and B. Scheers, ‘‘Passive radio system for real-time drone pp. 211–224.
detection and DoA estimation,’’ in Proc. Int. Conf. Mil. Commun. Inf. [204] M. Ezuma, F. Erden, C. K. Anjinappa, O. Ozdemir, and I. Guvenc,
Syst. (ICMCIS), Warsaw, Poland, May 2018, pp. 1–6. ‘‘Micro-UAV detection and classification from RF fingerprints using
[181] R. K. Miranda, D. A. Ando, J. P. C. L. da Costa, and M. T. de Oliveira, machine learning techniques,’’ in Proc. IEEE Aerosp. Conf., Big Sky, MT,
‘‘Enhanced direction of arrival estimation via received signal strength USA, Mar. 2019, pp. 1–13.
of directional antennas,’’ in Proc. IEEE Int. Symp. Signal Process. Inf. [205] T. Yucek and H. Arslan, ‘‘A survey of spectrum sensing algorithms for
Technol. (ISSPIT), Louisville, KY, USA, Dec. 2018, pp. 162–167. cognitive radio applications,’’ IEEE Commun. Surveys Tuts., vol. 11,
[182] I.-F. Kenmogne, V. Drevelle, and E. Marchand, ‘‘Image-based UAV no. 1, pp. 116–130, 1st Quart., 2009.
localization using interval methods,’’ in Proc. IEEE/RSJ Int. Conf. Intell. [206] P. Molchanov, K. Egiazarian, J. Astola, R. I. A. Harmanny, and
Robots Syst. (IROS), Vancouver, BC, Canada, Sep. 2017, pp. 5285–5291. J. J. M. de Wit, ‘‘Classification of small UAVs and birds by micro-Doppler
[183] M. Vrba, D. Heřt, and M. Saska, ‘‘Onboard marker-less detection and signatures,’’ in Proc. Eur. Radar Conf., Nuremberg, Germany, Oct. 2013,
localization of non-cooperating drones for their safe interception by an pp. 172–175.
autonomous aerial system,’’ IEEE Robot. Autom. Lett., vol. 4, no. 4, [207] P. Molchanov, K. Egiazarian, J. Astola, A. Totsky, S. Leshchenko, and
pp. 3402–3409, Oct. 2019. M. P. Jarabo-Amores, ‘‘Classification of aircraft using micro-Doppler
[184] M. Vrba and M. Saska, ‘‘Marker-less micro aerial vehicle detection and bicoherence-based features,’’ IEEE Trans. Aerosp. Electron. Syst., vol. 50,
localization using convolutional neural networks,’’ IEEE Robot. Autom. no. 2, pp. 1455–1467, Apr. 2014.
Lett., vol. 5, no. 2, pp. 2459–2466, Apr. 2020. [208] B. Torvik, K. E. Olsen, and H. Griffiths, ‘‘Classification of birds and
[185] F. Gökçe, G. Üçoluk, E. Şahin, and S. Kalkan, ‘‘Vision-based detection UAVs based on radar polarimetry,’’ IEEE Geosci. Remote Sens. Lett.,
and distance estimation of micro unmanned aerial vehicles,’’ Sensors, vol. 13, no. 9, pp. 1305–1309, Sep. 2016.
vol. 15, no. 9, pp. 23805–23846, Sep. 2015. [209] J. Ren and X. Jiang, ‘‘Regularized 2-D complex-log spectral analysis
[186] F. Hoffmann, M. Ritchie, F. Fioranelli, A. Charlish, and H. Griffiths, and subspace reliability analysis of micro-Doppler signature for UAV
‘‘Micro-Doppler based detection and tracking of UAVs with multistatic detection,’’ Pattern Recognit., vol. 69, pp. 225–237, Sep. 2017.
radar,’’ in Proc. IEEE Radar Conf. (RadarConf), Philadelphia, PA, USA, [210] B. K. Kim, H.-S. Kang, and S.-O. Park, ‘‘Drone classification using con-
May 2016, pp. 1–6. volutional neural networks with merged Doppler images,’’ IEEE Geosci.
[187] U.S. Federal Aviation Administration. (2019). Unmanned Aircraft Remote Sens. Lett., vol. 14, no. 1, pp. 38–42, Jan. 2017.
System Detection—Technical Considerations. [Online]. Available: [211] B. R. Mahafza, Radar Systems Analysis and Design Using MATLAB.
https://www.faa.gov/airports/airport_safety/media/Attachment-3-UAS- Boca Raton, FL, USA: CRC Press, 2013.
Detection-Technical-Considerations.pdf [212] A. Stateczny and J. Lubczonek, ‘‘FMCW radar implementation in river
[188] D. He, Y. Qiao, S. Chan, and N. Guizani, ‘‘Flight security and safety information services in poland,’’ in Proc. 16th Int. Radar Symp. (IRS),
of drones in airborne fog computing systems,’’ IEEE Commun. Mag., Dresden, Germany, Jun. 2015, pp. 852–857.
vol. 56, no. 5, pp. 66–71, May 2018. [213] J. Farlik, M. Kratky, J. Casar, and V. Stary, ‘‘Multispectral detection of
[189] Y. Mao, C. You, J. Zhang, K. Huang, and K. B. Letaief, ‘‘A survey on commercial unmanned aerial vehicles,’’ Sensors, vol. 19, no. 7, p. 1517,
mobile edge computing: The communication perspective,’’ IEEE Com- Mar. 2019.
mun. Surveys Tuts., vol. 19, no. 4, pp. 2322–2358, 4th Quat., 2017. [214] N. Eriksson, ‘‘Conceptual study of a future drone detection system—
[190] DJI Support. (Aug. 2017). How to Use DJI’s Return to Home (RTH) Countering a threat posed by a disruptive technology,’’ M.S. thesis,
Safely. [Online]. Available: https://store.dji.com/guides/how-to-use-the- Chalmers Univ. Technol., Gothenburg, Sweden, 2018.
djis-return-to-home/ [215] V. C. Chen, The Micro-Doppler Effect in Radar. Boston, MA, USA:
[191] RoboTiCan. (Apr. 2020). Goshawk. [Online]. Available: Artech House, 2019.
https://robotican.net/goshawk/ [216] B. K. Kim, H.-S. Kang, and S.-O. Park, ‘‘Experimental analysis of small
[192] L. Hauzenberger and E. H. Ohlsson, ‘‘Drone detection using audio anal- drone polarimetry based on micro-Doppler signature,’’ IEEE Geosci.
ysis,’’ M.S. thesis, Lund Univ., Lund, Sweden, Jun. 2015. Remote Sens. Lett., vol. 14, no. 10, pp. 1670–1674, Oct. 2017.
[193] Droneshield. (Apr. 2020). Dronesentry. [Online]. Available: [217] R. Guay, G. Drolet, and J. R. Bray, ‘‘Measurement and modelling of the
https://www.droneshield.com/sentry dynamic radar cross-section of an unmanned aerial vehicle,’’ IET Radar,
[194] Alsok. (Apr. 2020). [Online]. Available: https://www.alsok.co.jp/en/ Sonar Navigat., vol. 11, no. 7, pp. 1155–1160, Jul. 2017.
[195] B. Harvey and S. O’Young, ‘‘Acoustic detection of a fixed-wing UAV,’’ [218] V. C. Chen, F. Li, S. S. Ho, and H. Wechsler, ‘‘Analysis of micro-Doppler
Drones, vol. 2, no. 1, pp. 4–22, Jan. 2018. signatures,’’ IEE Proc.-Radar, Sonar Navigat., vol. 150, no. 4, p. 271,
[196] G. Ottoy and L. De Strycker, ‘‘An improved 2D triangulation algo- Nov. 2003.
rithm for use with linear arrays,’’ IEEE Sensors J., vol. 16, no. 23, [219] V. C. Chen, F. Li, S.-S. Ho, and H. Wechsler, ‘‘Micro-Doppler effect in
pp. 8238–8243, Dec. 2016. radar: Phenomenon, model, and simulation study,’’ IEEE Trans. Aerosp.
[197] E. E. Case, A. M. Zelnio, and B. D. Rigling, ‘‘Low-cost acoustic array for Electron. Syst., vol. 42, no. 1, pp. 2–21, Jan. 2006.
small UAV detection and tracking,’’ in Proc. IEEE Nat. Aerosp. Electron. [220] V. C. Chen and S. Qian, ‘‘Joint time-frequency transform for radar range-
Conf., Dayton, OH, USA, Jul. 2008, pp. 110–113. Doppler imaging,’’ IEEE Trans. Aerosp. Electron. Syst., vol. 34, no. 2,
[198] A. Yakubovskiy, H. Salloum, A. Sutin, A. Sedunov, N. Sedunov, and pp. 486–499, Apr. 1998.
D. Masters, ‘‘Feature extraction for acoustic classification of small air- [221] V. C. Chen, D. Tahmoush, and W. J. Miceli, Radar Micro-Doppler
craft,’’ in Proc. IEEE Workshop Appl. Signal Process. Audio Acoust. Signature: Processing and Applications. London, U.K.: The Institution
(WASPAA), New Paltz, NY, USA, Oct. 2015, pp. 1–5. of Engineering and Technology, 2014.
[199] A. Bernardini, F. Mangiatordi, E. Pallotti, and L. Capodiferro, [222] J. J. M. de Wit, R. I. A. Harmanny, and G. Prémel-Cabic, ‘‘Micro-Doppler
‘‘Drone detection by acoustic signature identification,’’ Electron. Imag., analysis of small UAVs,’’ in Proc. Int. Eur. Radar Conf., Amsterdam,
vol. 2017, no. 10, pp. 60–64, Jan. 2017. The Netherlands, Feb. 2012, pp. 210–213.
[200] J. Kim and D. Kim, ‘‘Neural network based real-time UAV detection and [223] J. J. M. de Wit, R. I. A. Harmanny, and P. Molchanov, ‘‘Radar micro-
analysis by sound,’’ J. Adv. Inf. Technol. Converg., vol. 8, no. 1, pp. 43–52, Doppler feature extraction using the singular value decomposition,’’ in
Jul. 2018. Proc. Int. Radar Conf., Lille, France, Oct. 2014, pp. 1–6.

168706 VOLUME 8, 2020

H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

[224] S. Rahman and D. A. Robertson, ‘‘Millimeter-wave micro-Doppler mea- [246] A. Li, Q. Wu, and R. Zhang, ‘‘UAV-enabled cooperative jamming for
surements of small UAVs,’’ in Proc. SPIE, vol. 10188, May 2017, improving secrecy of ground wiretap channel,’’ IEEE Wireless Commun.
Art. no. 101880T. Lett., vol. 8, no. 1, pp. 181–184, Feb. 2019.
[225] R. J. Fontana, E. A. Richley, A. J. Marzullo, L. C. Beard, R. W. T. Mulloy, [247] J. Noh, Y. Kwon, Y. Son, H. Shin, D. Kim, J. Choi, and Y. Kim,
and E. J. Knight, ‘‘An ultra wideband radar for micro air vehicle appli- ‘‘Tractor beam: Safe-hijacking of consumer drones with adaptive GPS
cations,’’ in Proc. IEEE Conf. Ultra Wideband Syst. Technol., Baltimore, spoofing,’’ ACM Trans. Privacy Secur., vol. 22, no. 2, pp. 12:1–12:26,
MD, USA, May 2002, pp. 187–191. Apr. 2019.
[226] T. Mizushima, R. Nakamura, and H. Hadama, ‘‘Reflection character- [248] A. J. Kerns, D. P. Shepard, J. A. Bhatti, and T. E. Humphreys, ‘‘Unmanned
istics of ultra-wideband radar echoes from various drones in flight,’’ aircraft capture and control via GPS spoofing,’’ J. Field Robot., vol. 31,
in Proc. IEEE Topical Conf. Wireless Sensors Sensor Netw. (WiSNeT), no. 4, pp. 617–636, Apr. 2014.
San Antonio, TX, USA, Jan. 2020. [249] M. Hooper, Y. Tian, R. Zhou, B. Cao, A. P. Lauf, L. Watkins,
[227] C. J. Li and H. Ling, ‘‘An investigation on the radar signatures of W. H. Robinson, and W. Alexis, ‘‘Securing commercial WiFi-based
small consumer drones,’’ IEEE Antennas Wireless Propag. Lett., vol. 16, UAVs from common security attacks,’’ in Proc. IEEE Mil. Commun. Conf.
pp. 649–652, Jul. 2017. (MILCOM), Baltimore, MD, USA, Nov. 2016, pp. 1213–1218.
[228] B. Torvik, A. Knapskog, O. Lie-Svendsen, K. E. Olsen, and [250] N. Summers. (Oct. 2016). Icarus Machine Can Commandeer a Drone
H. D. Griffiths, ‘‘Amplitude modulation on echoes from large birds,’’ in Mid-Flight. [Online]. Available: https://www.engadget.com/2016-10-28-
Proc. 11th Eur. Radar Conf., Rome, Italy, Oct. 2014, pp. 177–180. icarus-hijack-dmsx-drones.html
[229] I. Güvenç, O. Ozdemir, Y. Yapici, H. Mehrpouyan, and D. Matolak, [251] K. Moskvitch. (Feb. 2014). Are Drones the Next Target for Hackers?
‘‘Detection, localization, and tracking of unauthorized UAS and jam- [Online]. Available: https://www.bbc.com/future/article/20140206-can-
mers,’’ in Proc. IEEE/AIAA 36th Digit. Avionics Syst. Conf. (DASC), drones-be-hacked
St. Petersburg, FL, USA, Sep. 2017, pp. 1–10. [252] Y. Zeng, J. Lyu, and R. Zhang, ‘‘Cellular-connected UAV: Potential, chal-
[230] M. Ritchie, F. Fioranelli, H. Griffiths, and B. Torvik, ‘‘Monostatic lenges, and promising technologies,’’ IEEE Wireless Commun., vol. 26,
and bistatic radar measurements of birds and micro-drone,’’ in Proc. no. 1, pp. 120–127, Feb. 2019.
IEEE Radar Conf. (RadarConf), Philadelphia, PA, USA, May 2016, [253] W. A. Radasky, C. E. Baum, and M. W. Wik, ‘‘Introduction to the
pp. 1–5. special issue on high-power electromagnetics (HPEM) and intentional
[231] T. Müller, ‘‘Robust drone detection for day/night counter-UAV with electromagnetic interference (IEMI),’’ IEEE Trans. Electromagn. Com-
static VIS and SWIR cameras,’’ Proc. SPIE, vol. 10190, pp. 302–313, pat., vol. 46, no. 3, pp. 314–321, Aug. 2004.
May 2017. [254] Raytheon. (Apr. 2020). Phaser High-Power Microwave System. [Online].
[232] P. Andraši, T. Radišić, M. Muštra, and J. Ivošević, ‘‘Night-time detection Available: https://www.raytheon.com/capabilities/products/phaser-high-
of UAVs using thermal infrared camera,’’ Transp. Res. Procedia, vol. 28, power-microwave-system
pp. 183–190, Jan. 2017. [255] B. Zohuri, ‘‘High-power microwave energy as weapon,’’ in Directed-
[233] A. Rozantsev, V. Lepetit, and P. Fua, ‘‘Detecting flying objects using a Energy Beam Weapons. Cham, Switzerland: Springer, 2019, pp. 269–308,
single moving camera,’’ IEEE Trans. Pattern Anal. Mach. Intell., vol. 39, doi: 10.1007/978-3-030-20794-6_4.
no. 5, pp. 879–892, May 2017. [256] J. Lin and P. Singer. (Feb. 2017). Drones, Lasers, and Tanks: China Shows
[234] M. Saqib, S. D. Khan, N. Sharma, and M. Blumenstein, ‘‘A study on Off Its Latest Weapons. [Online]. Available: https://www.popsci.com/
detecting drones using deep convolutional neural networks,’’ in Proc. 14th china-new-weapons-lasers-drones-tanks/
IEEE Int. Conf. Adv. Video Signal Based Surveill. (AVSS), Lecce, Italy, [257] India Today. (Sep. 2015). KALI: India’s Weapon to Destroy Any Uninvited
Aug. 2017, pp. 1–5. Missiles and Aircrafts. [Online]. Available: https://www.indiatoday.
[235] C. Craye and S. Ardjoune, ‘‘Spatio-temporal semantic segmentation for in/education-today/gk-current-affairs/story/indias-top-secret-weapon-
drone detection,’’ in Proc. 16th IEEE Int. Conf. Adv. Video Signal Based 264111-2015-09-21
Surveill. (AVSS), Taipei, Taiwan, Sep. 2019, pp. 1–5. [258] D. Sudakov. (Aug. 2016). Russia’s Combat Laser Weapons Declassi-
[236] V. Magoulianitis, D. Ataloglou, A. Dimou, D. Zarpalas, and P. Daras, fied. [Online]. Available: https://www.pravdareport.com/russia/135198-
‘‘Does deep super-resolution enhance UAV detection?’’ in Proc. 16th russia_laser_weapons/
IEEE Int. Conf. Adv. Video Signal Based Surveill. (AVSS), Taipei, Taiwan, [259] MBDA Missile Systems. (Sep. 2017). Dragonfire Laser Turret
Sep. 2019, pp. 1–6. Unveiled at DSEI 2017. [Online]. Available: https://www.mbda-
[237] C. Cigla, R. Thakker, and L. Matthies, ‘‘Onboard stereo vision for drone systems.com/press-releases/dragonfire-laser-turret-unveiled-dsei-2017/
pursuit or sense and avoid,’’ in Proc. IEEE/CVF Conf. Comput. Vis. Pat- [260] Daily Sabah. (Sep. 2019). Turkey’s Laser Weapon ARMOL Passes
tern Recognit. Workshops (CVPRW), Salt Lake City, UT, USA, Jun. 2018, Acceptance Tests. [Online]. Available: https://www.dailysabah.
pp. 738–746. com/defense/2019/09/30/turkeys-laser-weapon-armol-passes-
[238] A. Schumann, L. Sommer, J. Klatte, T. Schuchert, and J. Beyerer, ‘‘Deep acceptance-tests
cross-domain flying object classification for robust UAV detection,’’ in [261] K. D. Atherton. (Aug. 2015). Boeing Unveils Its Anti-Drone Laser
Proc. 14th IEEE Int. Conf. Adv. Video Signal Based Surveill. (AVSS), Weapon. [Online]. Available: https://www.popsci.com/boeing-unveils-
Lecce, Italy, Aug. 2017, pp. 1–6. compact-anti-drone-laser/
[239] T. Ringwald, L. Sommer, A. Schumann, J. Beyerer, and R. Stiefelhagen, [262] Agence France-Presse in Beijing. (Nov. 2014). China
‘‘UAV-Net: A fast aerial vehicle detector for mobile platforms,’’ in Proc. Unveils Laser Drone Defence System. [Online]. Available:
IEEE/CVF Conf. Comput. Vis. Pattern Recognit. Workshops (CVPRW), https://www.theguardian.com/world/2014/nov/03/china-unveils-laser-
Long Beach, CA, USA, Jun. 2019, pp. 1–9. drone-defence-system
[240] C. Aker and S. Kalkan, ‘‘Using deep networks for drone detection,’’ in [263] Raytheon. (Apr. 2020). Forty-Five Down. [Online]. Available:
Proc. 14th IEEE Int. Conf. Adv. Video Signal Based Surveill. (AVSS), https://www.raytheon.com/news/feature/forty-five-down
Lecce, Italy, Aug. 2017, pp. 1–6. [264] D. Gettinger and A. H. Michel. (Jul. 2014). A Brief History
[241] M. Hammer, M. Hebel, B. Borgmann, M. Laurenzis, and M. Arens, of Hamas and Hezbollah’s Drones. [Online]. Available:
‘‘Potential of LiDAR sensors for the detection of UAVs,’’ Proc. SPIE, https://dronecenter.bard.edu/hezbollah-hamas-drones/
vol. 10636, pp. 39–45, May 2018. [265] Smart Rounds, Inc. (Oct. 2019). SAVAGE—Smart Anti-Drone Weapon.
[242] B. Kim, D. Khan, C. Bohak, W. Choi, H. Lee, and M. Kim, ‘‘V-RBNN [Online]. Available: https://www.prnewswire.com/news-releases/savage-
based small drone detection in augmented datasets for 3D LADAR sys- smart-anti-drone-weapon-300941541.html
tem,’’ Sensors, vol. 18, no. 11, p. 3825, Nov. 2018. [266] (2020). AerialX: DroneBullet. [Online]. Available: https://www.
[243] S. Rahman and D. A. Robertson, ‘‘Radar micro-Doppler signatures of aerialx.com/defeat.shtml
drones and birds at K-band and W-band,’’ Sci. Rep., vol. 8, no. 1, pp. 1–11, [267] G. Olivares, L. Gomez, J. E. de los Monteros, R. J. Baldridge,
Nov. 2018. C. Zinzuwadia, and T. Aldag, ‘‘Volume II—UAS airborne collision sever-
[244] W. Xu, K. Ma, W. Trappe, and Y. Zhang, ‘‘Jamming sensor networks: ity evaluation—Quadcopter,’’ Nat. Inst. Aviation Res., Wichita, KS, USA,
Attack and defense strategies,’’ IEEE Netw., vol. 20, no. 3, pp. 41–47, Tech. Rep. DOT/FAA/AR xx/xx, Jul. 2017.
May/Jun. 2006. [268] Ban Lethal. (Apr. 2020). Slaugtherbots. [Online]. Available:
[245] A. Mpitziopoulos, D. Gavalas, C. Konstantopoulos, and G. Pantziou, https://autonomousweapons.org/
‘‘A survey on jamming attacks and countermeasures in WSNs,’’ IEEE [269] Anduril Industries. (Apr. 2020). Anvil. [Online]. Available:
Commun. Surveys Tuts., vol. 11, no. 4, pp. 42–56, 4th Quat. 2009. https://www.anduril.com/

VOLUME 8, 2020 168707

 H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

[270] Raytheon. (Apr. 2020). Coyote. [Online]. Available: https://www. [298] Zion Market Research. (Feb. 2019). Anti-Drone Market by System,
raytheon.com/capabilities/products/coyote by Technology, and by End-User: Global Industry Perspective,
[271] CDET. (Apr. 2020). RAM UAV. [Online]. Available: https://ramuav.com/ Comprehensive Analysis, and Forecast, 2016–2025. [Online]. Available:
[272] Chenega. (Apr. 2020). Counter-UAV Solutions. [Online]. Available: https://www.zionmarketresearch.com/market-analysis/anti-drone-
https://www.chenegaeurope.com/media/1221/counteruav.pdf market
[273] Drone Defence. (Apr. 2020). NetGun X1. [Online]. Available: [299] Modor Intelligence. (2019). Anti-Drone Market—Growth, Trends,
https://www.dronedefence.co.uk/products/netgun-x1/ and Forecast (2020–2025). [Online]. Available: https://www.
[274] Droptec. (Apr. 2020). Dropster Net Gun. [Online]. Available: mordorintelligence.com/industry-reports/anti-drone-market
https://www.droptec.ch/product [300] A. Charlton. (Dec. 2019). Forbes: Drone (Regulation) Wars: U.S.
[275] UAVOS. (Apr. 2020). Interception System for Small Sized Unmanned
and E.U. Face Off. [Online]. Available: https://www.forbes.com/
Vehicles. [Online]. Available: https://www.uavos.com/products/uas-
sites/andrewcharlton5/2019/12/09/drone-regulation-warsus-and-eu-face-
payloads/interception-system-for-uav
off/#5dd37c343412
[276] ALS Defense. (Apr. 2020). SKYNET Mi-5. [Online]. Available:
https://www.lesslethal.com/products/12-gauge/als12skymi-5-detail [301] T. McMullan. (Mar. 2019). How Swarming Drones Will Change
[277] Delft Dynamics. (Apr. 2020). Dronecatcher. [Online]. Available: Warfare. [Online]. Available: https://www.bbc.com/news/technology-
https://dronecatcher.nl/ 47555588
[278] Fortem Technologies. (Apr. 2020). Drone Hunter. [Online]. Available: [302] J. Spero. (Jun. 2019). Concerns Rise Over Use of Drones in a Swarm
https://fortemtech.com/products/dronehunter/ Attack. [Online]. Available: https://www.ft.com/content/c51fa3f8-8d10-
[279] Science Technology. (Apr. 2020). Aeroguard. [Online]. Available: 11e9-a1c1-51bf8f989972
https://www.sci.com/aeroguard/ [303] C. Scotti. (Aug. 2016). Who Can be Killed by a Drone?
[280] Search Systems. (Apr. 2020). Sparrowhawk. [Online]. Available: US Reveals the Rules of Engagement. [Online]. Available:
http://www.searchsystems.eu/sparrowhawk.html http://www.thefiscaltimes.com/2016/08/09/Who-Can-Be-Killed-Drone-
[281] SKYLOCK. (Apr. 2020). Counter Drone Net Catcher. [Online]. Avail- US-Reveals-Rules-Engagement
able: https://www.skylock1.com/counter-drone-systems/ [304] Aveillant. Accessed: Sep. 12, 2020. [Online]. Available:
[282] D. Sathyamoorthy, ‘‘A review of security threats of unmanned aerial https://www.aveillant.com/
vehicles and mitigation steps,’’ J. Defence Secur., vol. 6, no. 1, pp. 81–97, [305] Blighter. Accessed: Sep. 12, 2020. [Online]. Available:
Oct. 2015. https://www.blighter.com/
[283] E. Ackerman. (Apr. 2015). South Korea Prepares for Drone
[306] Drone Citadel. Accessed: Sep. 12, 2020. [Online]. Available:
vs. Drone Combat. [Online]. Available: https://spectrum.ieee.
https://dronecitadel.com/
org/automaton/robotics/drones/south-korea-drone-vs-drone
[284] OpenWorks Engineering. (Apr. 2020). Skywall. [Online]. Available: [307] (2020). Dronedefence. [Online]. Available: https://www.dronedefence.
https://openworksengineering.com/skywall-patrol/ co.uk/
[285] K. D. Atherton. (Feb. 2016). Trained Police Eagles Attack Drones on [308] DroneShield. Accessed: Sep. 12, 2020. [Online]. Available:
Command. [Online]. Available: https://www.popsci.com/eagles-attack- https://www.droneshield.com/
drones-at-police-command/ [309] Liteye. Accessed: Sep. 12, 2020. [Online]. Available: https://liteye.com
[286] A. Y. Javaid, W. Sun, V. K. Devabhaktuni, and M. Alam, ‘‘Cyber security [310] Lockheed Martin. Accessed: Sep. 12, 2020. [Online]. Available:
threat analysis and modeling of an unmanned aerial vehicle system,’’ in https://www.lockheedmartin.com/en-us/index.html
Proc. IEEE Conf. Technol. Homeland Secur. (HST), Waltham, MA, USA, [311] Northrop Grumman. Accessed: Sep. 12, 2020. [Online]. Available:
Nov. 2012, pp. 585–590. https://www.
[287] C. G. L. Krishna and R. R. Murphy, ‘‘A review on cybersecurity vulner- northropgrumman.com/
abilities for unmanned aerial vehicles,’’ in Proc. IEEE Int. Symp. Saf., [312] Raytheon. Accessed: Sep. 12, 2020. [Online]. Available:
Secur. Rescue Robot. (SSRR), Shanghai, China, Oct. 2017, pp. 194–199. https://www.raytheon.com
[288] (Apr. 2020). Pixhawk. [Online]. Available: https://pixhawk.org/
[313] Saab AB. Accessed: Sep. 12, 2020. [Online]. Available:
[289] Our Bureau. (Dec. 2014). Us Navy Laser Weapon Fires at
https://saabgroup.com/
$1 Per Shot. [Online]. Available: https://www.defenseworld.
net/news/11684/US_Navy_Laser_Weapon_Fires_at__1_Per_Shot#. [314] SkySafe. Accessed: Sep. 12, 2020. [Online]. Available:
XrdHc2gzaUl https://www.skysafe.io/
[290] Fortune Business Insights. (Feb. 2020). Commercial Drones Mar- [315] Thales. Accessed: Sep. 12, 2020. [Online]. Available:
ket Size, Share & Industry Analysis, by Product, by Technology, by https://www.thalesgroup.com/en
System, by Industry, and Regional Forecast, 2019–2026. [Online]. [316] WhiteFox Defense Technologies. Accessed: Sep. 12, 2020. [Online].
Available: https://www.fortunebusinessinsights.com/commercial-drone- Available: https://www.whitefoxdefense.com/
market-102171 [317] Lockheed Martin. Accessed: Sep. 12, 2020. [Online]. Available:
[291] K. Wackwitz. (Dec. 2019). The Counter-Drone Market Report 2020. https://www.lockheedmartin.com/en-us/products/indago-vtol-
[Online]. Available: https://www.droneii.com/project/counter-drone- uav.html
market-report-2020 [318] Northrop: Globalhawk. Accessed: Sep. 12, 2020. [Online]. Available:
[292] Grand View Research. (May 2019). Anti-Drone Market Size Worth $4.5 https://www.northropgrumman.com/air/globalhawk/
Billion by 2026. [Online]. Available: https://www.grandviewresearch. [319] Raytheon: Silverfox. Accessed: Sep. 12, 2020. [Online]. Available:
com/press-release/global-anti-drone-market https://www.raytheon.com/capabilities/products/silverfox
[293] Research and Markets. (Sep. 2019). Global Counter-UAS
Market: Focus on Technology, Application, End Users—Analysis [320] Lockheed Martin: Drone Swarms. Accessed: Sep. 12, 2020.
and Forecast, 2019–2024. [Online]. Available: https://www. [Online]. Available: https://www.lockheedmartin.com/en-
researchandmarkets.com/reports/4845575/global-counter-uas-anti- us/news/features/2016/webt-laser-swarms-drones.html
drone-market-focus-on [321] (2020). Northrop Grumman: Mobile Application for UAS Identification
[294] Homeland Security Market Research. (Feb. 2019). Anti-Drone (MAUI). [Online]. Available: https://news.northropgrumman.
Market & Technologies—2019-2023. [Online]. Available: https:// com/news/releases/northrop-grumman-demonstrates-counter-uas-
homelandsecurityresearch.com/reports/counter-drone-market/ technologies-at-black-dart-exercise
[295] (2020). Droneii: Drone Industry Insight. [Online]. Available: [322] Leonardo DRS Press Release. (Oct. 2017). U.S. Army
https://www.droneii.com/ Awards Leonardo DRS Contract for Production of Counter-
[296] Markets and Markets. (Nov. 2019). Anti-Drone Market by Tech- Drone Capability. [Online]. Available: https://www.
nology, Application, Vertical, and Geography—Global Forecast to verticalmag.com/press-releases/u-s-army-awards-leonardo-drs-contract-
2024. [Online]. Available: https://www.marketsandmarkets.com/Market- production-counter-drone-capability/
Reports/anti-drone-market-177013645.html [323] S. Lewis. (Feb. 2020). Drone Defence Releases Solar Sentinel
[297] S. Kanowitz. (Dec. 2019). DOD Invests in Counter-Drone Technolo- Drone Detection System. [Online]. Available: https://www.
gies. [Online]. Available: https://gcn.com/articles/2019/12/11/counter- commercialdroneprofessional.com/drone-defence-releases-solar-
uas.aspx sentinel-drone-detection-system/

168708 VOLUME 8, 2020

H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

[324] Liteye Press Release. (Mar. 2020). Liteye & Citadel Push the Enve- HONGGU KANG (Student Member, IEEE)
lope for State-of-the-Art in Countering UAS Threats. [Online]. Avail- received the B.Sc. degree (summa cum laude) in
able: https://liteye.com/liteye-citadel-push-the-envelope-for-state-of-the- electronic engineering from Hanyang University,
art-in-countering-uas-threats/ Seoul, South Korea, in 2017, and the M.Sc. degree
[325] C. Lee. (Dec. 2019). Web Exclusive: Counter-UAS Company from the School of Electrical Engineering, Korea
Purchases Anti-Drone Shoulder Rifle. [Online]. Available: Advanced Institute of Science and Technology
https://www.nationaldefensemagazine.org/articles/2019/12/19/counter-
(KAIST), Daejeon, South Korea, in 2019, where
uas-company-purchases-anti-drone-shoulder-rifle
he is currently pursuing the Ph.D. degree. His
[326] MFIX Press Release. (Nov. 2018). Liteye and Northrop Grumman
Demonstrate Mobile, Networked, Electronic and Kinetic Capabilities research interests include signal processing for
to Counter Unmanned Threat Systems During Army’s Maneuver wireless communications, unmanned aerial vehi-
and Fires Integration Exercise (MFIX 18). [Online]. Available: cle communications, and machine learning. He was a recipient of the Korean
https://liteye.com/liteye-and-northrop-grumman-demonstrate-mobile- Institute of Communications and Information Sciences (KICS) Fall Sympo-
networked-electronic-and-kinetic-capabilities-to-counter-unmanned- sium Best Paper Award, in 2019.
threat-systems-during-armys-maneuver-and-fires-integration-exercise-
mfix-18/
[327] Liteye Press Release. (Apr. 2020). Liteye Expands their Counter
UAS Layered Approach With Raytheon Missiles & Defense’s Phaser.
[Online]. Available: https://liteye.com/liteye-expands-their-counter-uas-
layered-approach-with-raytheon-missiles-defenses-phaser/
[328] Paris La Défense. (Nov. 2017). Thales Completes the Acquisition of
Aveillant, World Pioneer in Holographic Radar Technology. [Online].
JINGON JOUNG (Senior Member, IEEE)
Available: https://www.aveillant.com/thales-completes-the-acquisition-
of-aveillant-world-pioneer-in-holographic-radar-technology/
received the B.S. degree in radio communication
[329] B. Stevenson. (Apr. 2016). Monaco to Operate Counter-UAV System. engineering from Yonsei University, Seoul, South
[Online]. Available: https://www.flightglobal.com/civil-uavs/monaco-to- Korea, in 2001, and the M.S. and Ph.D. degrees in
operate-counter-uav-system/120293.article electrical engineering and computer science from
[330] A. Chua. (Aug. 2017). Singapore Acquires Radar System Able KAIST, Daejeon, South Korea, in 2003 and 2007,
to Spot Small Drones up to 5km Away. [Online]. Available: respectively.
https://www.todayonline.com/singapore/spore-acquires-radar-system- He was a Postdoctoral Fellow with KAIST
able-spot-small-drones-5km-away and the University of California at Los Angeles
[331] Airport Technology. (Jul. 2017). Aveillant Installs Drone Detection (UCLA), CA, USA, in 2007 and 2008, respec-
Radar at Paris Charles de Gaulle Airport. [Online]. Available: tively. He was a Scientist with the Institute for Infocomm Research (I2 R),
https://www.airport-technology.com/news/newsaveillant-installs-first- Agency for Science, Technology and Research (A*STAR), Singapore, from
airport-drone-detection-radar-at-charles-de-gaulle-5863816/
2009 to 2015. He joined Chung-Ang University (CAU), Seoul, in 2016,
[332] DroneShield. (May 2019). Thales Purchases Droneshield Limited
as a Faculty Member. He is currently an Associate Professor with the
Solutions, Aims for Integration With Existing Technologies.
[Online]. Available: https://www.droneshield.com/all-press- School of Electrical and Electronics Engineering, CAU, where he is also
coverage/2019/5/1/thales-purchases-droneshield-limited-solutions- the Principal Investigator of the Intelligent Wireless Systems Laboratory.
aims-for-integration-with-existing-technologies His research interests include communication signal processing, numerical
[333] O. Sami Oubbati, M. Atiquzzaman, T. Ahamed Ahanger, and A. Ibrahim, analysis, algorithms, and machine learning.
‘‘Softwarization of UAV networks: A survey of applications and future Dr. Joung was a recipient of the First Prize of the Intel-ITRC Student Paper
trends,’’ IEEE Access, vol. 8, pp. 98073–98125, 2020. Contest, in 2006. He was recognized as an Exemplary Reviewer of the IEEE
[334] J. McCoy and D. B. Rawat, ‘‘Software-defined networking for unmanned COMMUNICATIONS LETTERS, in 2012, and the IEEE WIRELESS COMMUNICATIONS
aerial vehicular networking and security: A survey,’’ Electronics, vol. 8, LETTERS, in 2012, 2013, 2014, and 2019. He served as a Guest Editor for IEEE
no. 12, p. 1468, Dec. 2019. ACCESS, in 2016, and Electronics (MDPI), in 2019. He served on the Editorial
[335] G. Secinti, P. B. Darian, B. Canberk, and K. R. Chowdhury, ‘‘SDNs in Board for the APSIPA Transactions on Signal and Information Processing,
the sky: Robust end-to-end connectivity for aerial vehicular networks,’’ from 2014 to 2019. He is also serving as an Associate Editor for the IEEE
IEEE Commun. Mag., vol. 56, no. 1, pp. 16–21, Jan. 2018.
TRANSACTIONS ON VEHICULAR TECHNOLOGY and Sensors (MDPI).
[336] K. Chen, S. Zhao, N. Lv, W. Gao, X. Wang, and X. Zou, ‘‘Segment rout-
ing based traffic scheduling for the software-defined airborne backbone
network,’’ IEEE Access, vol. 7, pp. 106162–106178, 2019.
[337] B. Nogales, V. Sanchez-Aguero, I. Vidal, and F. Valera, ‘‘Adaptable and
automated small UAV deployments via virtualization,’’ Sensors, vol. 18,
no. 12, p. 4116, Nov. 2018.
[338] S. Sezer, S. Scott-Hayward, P. Chouhan, B. Fraser, D. Lake, J. Finnegan,
N. Viljoen, M. Miller, and N. Rao, ‘‘Are we ready for SDN? Implemen- JINYOUNG KIM (Member, IEEE) received the
tation challenges for software-defined networks,’’ IEEE Commun. Mag., B.S. degree in electrical engineering from KAIST,
vol. 51, no. 7, pp. 36–43, Jul. 2013. Daejeon, South Korea, in 2001, and the Ph.D.
[339] R. Amin, M. Reisslein, and N. Shah, ‘‘Hybrid SDN networks: A survey degree in business management from the Nanyang
of existing approaches,’’ IEEE Commun. Surveys Tuts., vol. 20, no. 4, Business School, Nanyang Technological Univer-
pp. 3259–3306, 4th Quart., 2018. sity, Singapore, in 2017.
[340] B. Han, V. Gopalakrishnan, L. Ji, and S. Lee, ‘‘Network function virtu- She attended the Technology and Policy Pro-
alization: Challenges and opportunities for innovations,’’ IEEE Commun. gram of the Engineering Systems Division, Mas-
Mag., vol. 53, no. 2, pp. 90–97, Feb. 2015. sachusetts Institute of Technology (MIT), USA,
[341] X. Foukas, G. Patounas, A. Elmokashfi, and M. K. Marina, ‘‘Network
from 2002 to 2005. She was an Assistant Manager
slicing in 5G: Survey and challenges,’’ IEEE Commun. Mag., vol. 55,
no. 5, pp. 94–100, May 2017. at Samsung Electronics, Suwon, South Korea, from 2005 to 2008. She was
[342] B. Deng, C. Jiang, H. Yao, S. Guo, and S. Zhao, ‘‘The next generation also an In-House Startup Mentor with the Seoul Global Startup Center, from
heterogeneous satellite communication networks: Integration of resource 2016 to 2017. She was a Lecturer with Dongguk University, from 2017 to
management and deep reinforcement learning,’’ IEEE Wireless Commun., 2019, and Chung-Ang University, in 2018. She has been a Research Professor
vol. 27, no. 2, pp. 105–111, Apr. 2020. with the Korea University Business School, Seoul, South Korea, since 2017.
[343] T. Hong, W. Zhao, R. Liu, and M. Kadoch, ‘‘Space-air-ground IoT Her research interests include entrepreneurship, technology innovation, and
network and related key technologies,’’ IEEE Wireless Commun., vol. 27, decision-making process under uncertainty. She is a member of the Decision
no. 2, pp. 96–104, Apr. 2020. Science Institute and Academy of Management.

VOLUME 8, 2020 168709

 H. Kang et al.: Protect Your Sky: A Survey of Counter UAV Systems

JOONHYUK KANG (Member, IEEE) received the YONG SOO CHO (Senior Member, IEEE) was
B.S.E. and M.S.E. degrees from Seoul National born in South Korea. He received the B.S. degree
University, Seoul, South Korea, in 1991 and 1993, in electronics engineering from Chung-Ang Uni-
respectively, and the Ph.D. degree in electrical versity, Seoul, South Korea, in 1984, the M.S.
and computer engineering from The University of degree in electronics engineering from Yonsei
Texas at Austin, Austin, in 2002. From 1993 to University, Seoul, in 1987, and the Ph.D. degree
1998, he was a Research Staff Member at Sam- in electrical and computer engineering from The
sung Electronics, Suwon, South Korea, where he University of Texas at Austin, Austin, TX, USA,
was involved in the development of DSP-based in 1991.
real-time control systems. In 2000, he was with During 1984, he was a Research Engineer at
Cwill Telecommunications, Austin, TX, USA, where he participated in Goldstar Electrical Company, Osan, South Korea. In 2001, he was a
the project for multicarrier CDMA systems with antenna array. He was Visiting Research Fellow with the Electronics and Telecommunications
a Visiting Scholar with the School of Engineering and Applied Sciences, Research Institute. Since 1992, he has been a Professor with the School
Harvard University, Cambridge, MA, USA, from 2008 to 2009. He is cur- of Electrical and Electronics Engineering, Chung-Ang University. He is the
rently a Faculty Member with the Department of Electrical Engineering author of 12 books, more than 400 conference and articles, and more than
(EE), KAIST, Daejeon, South Korea. His research interests include signal 120 patents. His research interests include the area of mobile communication
processing for cognitive radio, cooperative communication, physical-layer and digital signal processing, especially for MIMO OFDM and 5G.
security, and wireless localization. He is a member of the Korea Information Dr. Cho was a recipient of the Dr. Irwin Jacobs Award, in 2013. He served
and Communications Society and the Tau Beta Pi (the Engineering Honor as the President for the Korean Institute of Communications and Information
Society). He was a recipient of the Texas Telecommunication Consortium Sciences, in 2016.
Graduate Fellowship, from 2000 to 2002.

168710 VOLUME 8, 2020