2015 IEEE International Conference on Computer and Information Technology; Ubiquitous Computing and Communications;
Dependable, Autonomic and Secure Computing; Pervasive Intelligence and Computing
Survey on Clean Slate Cellular-IoT Standard
Proposals Wael Guibene, Keith E. Nolan, Mark Y. Kelly Intel Labs Europe, Intel, Leixlip, Dublin, Ireland wael.guibene@intel.com, keith.nolan@intel.com, mark.y.kelly@intel.com
AbstractIn this paper we investigate the proposals made by B. Characterising IoT
various industries for the Cellular Internet of Things (C-IoT). We start by introducing the context of C-IoT and demonstrate The IoT has become a term that encompasses a wide range how this technology is closely linked to the Low Power-Wide of new applications and services built on objects that can Area (LPWA) technologies and networks. An in-depth look and compute and communicate. As a result, it has become a term system level evaluation is given for each clean slate technology that has a degree of fuzziness; it cannot be dened in terms and a comparison is made based on its specications. Index TermsInternet of Things Standards, Clean Slate Pro- of a single technology, communication protocol, form factor, posals, Evaluation of C-IoT, Low Power Wide Area application or service. Atzori et al. [1] surmise that the reason of today apparent I. I NTRODUCTION fuzziness around this term is a consequence of the name The internet of things (IoT) is considered to be the third Internet of Things itself, which syntactically is composed of wave of information technology after internet and mobile two terms. The rst one pushes towards a network oriented communications. The Internet of Things concept was formally vision of IoT, while the second one moves the focus on generic proposed in 2005 by the International Telecommunication objects to be integrated into a common framework. Union (ITU) where they published ITU Internet Report 2005: Perara et al. [2] outline some characteristics of the IoT. Internet of Things. It was noted in this report that all objects These are: can exchange information via the networks actively. Intelligence - data generation and transformation into The IoT is included by the US National Intelligence Council knowledge (NIC) in the list of six Disruptive Civil Technologies with Architecture - heterogeneous event and time-driven in- potential impacts on US national power. NIC foresees that by frastructure 2025 Internet nodes may reside in everyday things e.g. food Complex system - granular system of objects each with packages, furniture, paper documents, and more. It highlights different levels of resources future opportunities that will arise, starting from the idea that Size considerations - it is predicted that there will be 50- popular demand combined with technology advances could 100 billion devices connected to the Internet by 2020 drive widespread diffusion of an Internet of Things (IoT) Time considerations - handle billions of parallel and that could, like the present Internet, contribute invaluably to simultaneous events economic development. The possible threats deriving from a Space considerations - objects can be deployed over wide widespread adoption of such a technology are also stressed. geographical areas Indeed, it is emphasized that to the extent that everyday objects Everything-as-a-service - efcient, scalable, versatile ap- become information security risks, the IoT could distribute plications and services those risks far more widely than the Internet has to date. A. Background C. Introducing Cellular-IoT IoT is underpinned by an infrastructure comprising con- Recently, 3GPP approved a new item in GERAN named nected devices. Machine communications is a very large Cellular Internet of Things (C-IoT) [3]to specically emphasis market that is growing rapidly with reports stating a compound on clean slate proposals on cellular system support for ultra annual growth rate (CAGR)of well over 20%. However it is low complexity and low throughput, low power Internet of also a fragmented market, with a diverse range of applications Things applications. and products. Cisco, Analysys Mason, Market Publishers, This new group was motivated by the signicant growth Forrester, McKinsey, Ericsson, ABI Research and Strategy of M2M communication and the need to support of multi- Analytics, have all produced market reports for this market years battery operating IoT devices for indoor/deep-indoor sector forecasting rapid growth and an ever increasing diversity scenarios. of IoT applications and services. Analysys Mason forecast that C-IoT face many challenges and 3GPP is looking to evaluate the fastest-growing M2M sectors will be utilities and smart new proposals for a new access system with regard to the grids, healthcare, and security. following aspects:
978-1-5090-0154-5/15 $31.00 2015 IEEE 1596
DOI 10.1109/CIT/IUCC/DASC/PICOM.2015.240 Improved indoor coverage: This applies for indoor/deep- work involving TVWS has created the foundation for the indoor scenarios where generally RF signals penetration development of real-world tests and trials. These are needed to is quite limited. Novel C-IoT proposed PHYs and tech- better understand how the techniques and systems developed nologies have to demonstrate coverage capabilities for over the past decade could play a useful role in deployed these scenarios systems. Somewhat similar trials being carried out involving Support for massive number of low throughput devices: TVWS and in some cases, M2M, are currently underway The target applications for the C-IoT will need to transmit wordwide. few bytes of data, yet the trafc to be supported is IoT deployments can leverage a combination of communi- huge, as the C-IoT is expected to cover 55% SOM cations protocols e.g. MQTT, CoAP and Thread. (shares of market) of the estimated 50 Billion to 250 In summary, IoT infrastructure now comprises: Billion connected devices by 2020. The support of these A plethora of communications technologies and protocols devices should be made possible using either the GSM A myriad of potential applications and services exiting infrastructure and spectrum or fragmented unused Volumes of connected IoT devices on a scale unlike spectrum in the licensed operator band. what we have experienced during previous waves of Delay sensitivity: Two types of data trafc is dened. The information technology. delay tolerant non critical one, where data transmission Vision Mobile predict that the top IoT platforms in 2020 latency is not critical, e.g reading a meter. Whereas, the will be cross-vertical. Sector-specic platforms will be niche second type of data is delay-critical and it is suggested or in decline. Therefore to progress to a cross-vertical IoT that a delay requirement of 4 seconds is appropriate approach, a platform that can support the development and for the UL measured from the trigger event to the deployment of new IoT applications and services and reduce packet being ready for transmission from the base station the commercialisation complexity yet support the ve key IoT towards the network (e.g alarms). characteristics is required. This leads us to the concept of the Ultra low cost: Devices operating under C-IoT should IoT elastic abstraction layer. remain very cheap from cost and operation perspectives. Low power consumption: Power consumption is a key requirement inherited by C-IoT from the LPWA tech- nologies and networks with a target battery life of 10 years. Network architecture: The study should identify a Core Network architecture, security framework and Radio Ac- cess Network-Core Network interface, and associated protocol stacks for M2M which are suitable for the M2M/IoT market in the 2017 and onwards timeframe II. S TATE OF T HE A RT ON LPWA T ECHNOLOGIES From a technical perspective, IoT now leverages a wide range of communications technologies and protocols com- bined with application/service requirements that range from low bandwidth- high latency to high bandwidth-high frequency acquisition and short range. Wireline technology examples include Ethernet, optical - Fig. 1. LPWA Target Characteristics bre, BACnet, LONWorks, X10, CEBus. Cellular communica- tions technologies include GPRS, EDGE, HSPA, LTE, LTE+. Figure (1), shows how LPWA technologies (SigFox, Mesh networking technologies include Zigbee, ZWave, Neul and LoRa) compare to classical cellular technologies Flexnet, Wireless HART. (3G/4G/5G) and to mesh technologies (ZigBee) in terms of Low-power wide-area technologies include SigFox, LoRA, latency, cost, power consumption and geographical coverage. Adaptrum, WEIGHTLESS, and IEEE 802.22. Powerline carrier technologies include PRIME and G3, and III. C LEAN S LATE P ROPOSALS FOR C ELLULAR -I OT wireless local/metropolitan area network technologies include WiFi and WiMAX. A. 3GPP LTE Rel-13 Machine Type Communication Sub-1 GHz spectrum segments have been proven popular In the standardization efforts, the 3GPP group introduced in for new wireless communications services since the transition its release 13, the Machine Type Communication (MTC) [5] in from analogue to digital television and subsequent band clear- LTE. MTC aims at reducing devices cost, power consumption, ance. Unused available spectrum segments in the UHF/VHF extending coverage and handling massive numbers of devices. bands, known as television white space (TVWS) have been More specically, the 3GPP consortium recommends the fol- the subject of over a decade of research. Early theoretical lowing in terms of MTC denition and specications :
1597 Introducing a new UE category for MTC operations supporting: Reduced UE bandwidth to 1.4 MHz in both Uplink and Downlink Reduced maximum transmit power with an inte- grated PA implementation at the level of the UE Reduced support for DL transmission modes Relaxation of the UE processing capabilities (re- Fig. 3. Semtech CSS DL/UL Frame duced support of CSI/CQI, restricted modulation schemes...) (Chirp Spread Spectrum) modulation for both uplink and Provide power consumption reduction for the UE cate- downlink. This modulation allows data communication and gory/type dened above, both in normal coverage and positioning from deep indoor to fast moving outdoor devices, enhanced coverage, to target ultra-long battery life at the cost of a lower spectral efciency compared to the Opt for half duplex (FDD/TDD) as they are particularly narrow band alternative with tunable datarates. benecial from device complexity and power consump- tion point of view B. Nokias Narrow-band LTE-M Proposal Inspired from the 3GPP recommendations, Nokia derived its LTE-M [6]proposal for C-IoT. Nokias proposal focused on the subGHz licensed spectrum, by designing LTE-M to t in refarming only one GSM channel, or make it deployable on a guard band of an existing LTE networks. In LTE-M, each channel is comprised of 12 subcarriers with subcarrier spacing of 15 kHz, so a total of 180 KHz occupied bandwidth using either existing or modied LTE channels. Some of the channels such as data channels already t into 180 kHz (equivalent to 1 physical resource block, or PRB, in LTE). Other channels will need some modications to t into Fig. 4. Localization Principle based on DToA 180 kHz. The narrow-band communication is using less than 180 KHz bandwidth for UL and DL. These 200 KHz or spectrum can be either a decommissioned GSM channel or any unoccupied licensed band owned by the operator in the subGHz spectrum. The proposed narrow band approach achieves 20dB improve- ment in link budget compared to GSM using only 17dBm (50mW) RF output power from the UE (device) at 122bits/sec Fig. 2. LTE-M DL/UL Frames and can be in increased up to 7.8kbits/sec. Fig 2) illustrates the UL and DL frames in the 180 KHz D. Neul/Huawei Narrow-band Proposal LTE-M proposal. In the DL frame, Channels are TDM to- gether where M-subframe combines multiple 1ms subframes. Neuls proposal is also based on 180 KHz UL/DL channels New PBCH and synchronization signals are introduced and bandwidth for C-IoT and targets deployment in GSM or LTE scheduling resolution is in subframes instead of PRB. LTE-M guard bands [8] is designed to support QPSK and 16-QAM. In the DL channel, each 180 kHz resource block is divided LTE-M operates in half-duplex mode with a maximum into 12 channels, spaced by 15 kHz. This modulation allows datarate of 200 kbps in both UL and DL with no multiband both TDMA and FDMA access schemes and is conceptually support of fallback modes. si;ilar to OFDM. Each Channel is individually modulated (BPSK, QPSK, 16QAM) and pulse-shaped to minimize spec- C. Semtechs Combined Narrow-band and CSS Proposal tral side-lobes. The PHY data-rate is between 375 bps and 36 Semtechs PHY layer for C-IoT combines the LoRa mod- kbps. ulation, already used in the unlicensed subGHz ISM bands, a For the UP channels, each 180 kHz resource block is split chirp spread spectrum modulation (CSS), and an ultra narrow- into 36 UL channels distant by 5 kHz allowing also FDMA band [7]. and TDMA multiple access schemes. Each UL channel is Semtechs CSS PHY proposal allows both communication individually modulated in (D)QPSK, (D)BPSK or GMSK and and ranging (localization). Ranging is achieved using a CSS pulseshaped to minimize inter-user interference.
1598 We clearly see that if LTE inspired proposals (Rel 13 and LTE-M) inherit OFDMA access schemes in TDD/FDD half/full duplex modes, the other proprietary PHY proposals tend to be full duplex. The table shows also see that all clean slate proposals meet the requirements in terms of occupied BW so they can all t in a GSM channel and/or an LTE guard band. The achievable data-rates with these proposals vary from few Bps to few Kbps in adaptive systems and 200Kbps for LTE-M but with a xed data-rate. V. C ONCLUSIONS AND F UTURE D IRECTIONS Fig. 5. Neul-Huawei DL/UL Channelization In this paper, we compared the different clean slate pro- posals for C-IoT. Many improvement and features are brought E. SigFox UNB Proposal by these proposals compared to the already proposed LPWA SigFox further enhanced its UNB solution for ISM bands technologies such as localization based on differential time in C-IoT by introducing the cooperative ultra narrow band (C- of arrival. The different proposals seem to be respecting the UNB) technology as a PHY layer for its proposal [9]. constraints on BW and low power consumption and thus can The UL channel modulation is D-BPSK and devices trans- replace the already existing LPWA technologies. mit at 24 dBm. The UL channel ts on 200 KHz GSM, Some improvements are yet to be expected as these tech- where the 200 KHz are divided in 160 Hz micro-channels. nologies serve only few bits to kbits per second, which The achievable datarate with such modulation is 160 bps. can be a real disadvantage especially for several network As far as the DL is concerned the data-rate is xed to 600bps monitoring operations or rmware update over the air. A with a 2GFSK modulation with BT=1.0 and a F=+/-800 Hz good compromise is to have proposals capable to achieve bits to hundreds of kbits per seconds data-rates allowing bi- directional, energy efcient and low power operations as well as capable to serving high volume of connected devices. R EFERENCES [1] L. Atzori, A. Iera, and G. Morabito, The Internet of Things: A survey, Computer Networks, vol. 54, no. 15, pp. 27872805, 2010.. [2] C. Perera, A. Zaslavsky, P. Christen, and D. Georgakopoulos, Con- textaware computing for the internet of things: A survey, IEEE Com- munications Surveys and Tutorials, vol. 16, no. 1, pp. 414454, 2014.. [3] 3GPP GERAN Documentation [online] http://www.3gpp.org/DynaReport/TDocExMtgGP-6531298.htm [4] Ericsson White paper : LTE release 13 Uen 284 23-8267, April 2015 [5] 3GPP, Standardization of Machine-type Communications V0.2.4 (2014-06) [6] Ratasuk, R., Mangalvedhe, N., Ghosh, A., Vejlgaard, B. (2014, Fig. 6. SigFox DL/UL Proposal September). Narrowband LTE-M System for M2M Communication. In Vehicular Technology Conference (VTC Fall), 2014 IEEE 80th (pp. 1-5). IEEE. Chicago [7] 3GPP TSG GERAN 65, GP-150075, Combined narrow-band and IV. C OMPARISON OF C-I OT P ROPOSALS Spread Spectrum physical layer coverage and capacity simulations, Table I is summarizing main characteristics of the different Semtech, Mar. 2015. [8] 3GPP TSG GERAN 65, GP-150075, New Study Item on Cellular clean slate C-IoT proposals. System Support for Ultra Low Complexity and Low Throughput Internet The different proposals are compared in terms of used BW, of Things, Neul, April. 2014. uplink and downlink data-rates and full/half duplex operating [9] 3GPP TSG GERAN 65, GP-150057, C-UNB technology for Cellular IoT - Physical Layer, SigFox Wireless, Mar. 2015. mode.
Technology Bandwidth UL datarte DL datarate Duplex
