Eng. Sci.

, 2022, 19, 279–284

Engineered Science
DOI: https://dx.doi.org/10.30919/es8d705

A Table-Based End to End Encryption Technique Without Key

Exchange
Aakash Bharadawaj Srinivasan,1 Hemalatha S2,* and Ramathmika2

Abstract
Communication has become an integral part of all beings—both living and non-living. In most cases, everyone likes to keep
their communication personal and secure. Additionally, high-level information is expected to be secure. One of the methods
to secure information is encryption, where the information is safeguarded so that it cannot be detected by anybody other
than the intended recipient. In an encryption method, key management is essential for proper communication. The exchange
of keys may lead to leakage, modification, or change of the keys during transmission. Even in public-key cryptography, the key
management problem exists. The proposed method aims to reduce this problem by eliminating key transfers during the
encryption-decryption process. The present study uses the Internet Protocol (IP) addresses of the sender and the receiver to
securely encrypt the message. The channel is tightly secured in the proposed algorithm as there is no key transfer during
encryption and decryption techniques. A table generated by a polynomial function that produces non-linear outputs is used
for producing the ciphertext. This ensures that the communication is secure against multiple man-in-the-middle attacks. The
complexity of the proposed algorithm is computed and is found to be Big O notation (O(n)).
Keywords: Encryption; Decryption; Key Exchange; Cipher; IP address.
Received: 08 November 2021; Revised: 11 February 2022; Accepted: 05 May 2022.
Article type: Research article.

1. Introduction message, can be fed to a supercomputer and brute force

Encryption and decryption are important in the secure techniques can break down the ciphertext within minutes. To
transmission of a message. Without proper cryptic algorithms avoid such a threat of exposure the key must be eliminated.
in place, the message can be easily intercepted and misused. A The algorithm used in the encryption process is equally
man-in-the-middle can attack using brute force and can easily important. A loosely encrypted message is equivalent to a
capture the message transmitted. Almost all the data simple transmission of plaintext without any security. Hence,
transmitted is sensitive in today’s world and anyone cannot in most encryption algorithms, public or private keys are used
afford to leak important information. Anything from a basic to strengthen the cryptic message. The key protects the
messaging app to a bank’s internal server is prone to cyber- transmitted message using various encryption techniques. But
attacks and hence the sensitive information must be dealt with unfortunately, even the keys are subjected to attacks. Once the
cautiously. key and the encrypted message are obtained, they can be fed
Encryption is the most sought-after method to ensure the into super processors for decryption.
secure transmission of information. Numerous encryption The method proposed in the present study, which does not
methods are in use today. Most of these techniques use a key involve any key exchange, is based on a table constructed at
to encrypt—which plays a major role in generating the cipher. both the sender’s and the receiver's sides. The message to be
Key management is understandably one of the biggest encrypted is mapped to the table and the character obtained is
problems in the cryptosystem. The key, along with the substituted to generate the cipher. Another round of encryption
1
is employed to reduce the length of the generated cipher.
Robert H. Smith School of Business, University of Maryland,
In the proposed methodology, the transmission of the key
College Park, Maryland 20742, USA.
2 Department of Computer Science and Engineering, Manipal
doesn’t occur and hence, the risks of using a key for public
and private key encryption techniques are eliminated. A
Institute of Technology, Manipal Academy of Higher Education,
detailed comparison of multiple published studies is carried
Manipal -576104, Udupi, Karnataka, India.
out and the similarities and differences between these
*Email: hema.shama@manipal.edu (H. Shama)

© Engineered Science Publisher LLC 2022 Eng. Sci., 2022, 19, 279-284 | 279
Research article Engineered Science

algorithms with the proposed algorithm are laid down. encryption and decryption process. This algorithm uses public
A conventional cryptographic method is usually and private keys for its encryption technique and the final data
symmetric—where a single key is used for both encryption is sent to the web application platform. A drawback of the
and decryption, or asymmetric—where a public key is used study is that the running time increases as the length of the text
for encryption and a private key is used for decryption. In both increases.
models, the keys are essential, i.e., there should be an Rivest Cipher 4 RC4+ and Variably Modified Permutation
occurrence of key exchange for the communication to be Composition (VMPC) algorithms have a three-pass protocol
effective. Most of the encryption algorithms found in the scheme that is adopted for key exchange.[7] The complexity of
literature study or being used currently in different the algorithm comprises the process of the Key Scheduling
applications require a key and hence, key management is one Algorithm (KSA) and the Pseudo-Random Generation
of the major concerns to the researchers. Algorithm (PRGA). The symmetric algorithm proposed in the
Encryption is extensively used in almost all domains of study cannot be signed and hence it is prone to a man-in-the-
technology, such as networking, e-commerce, banking, and middle attack. A fast cryptosystem with identical encryption
database management. In some cases, the encrypted data and decryption process employs one keystream generator,
transmission is assessed using the User Datagram Protocol plaintext operations, and 180-degree matrix rotation.[8] This
(UDP) and Transmission Control Protocol (TCP).[1] TCP proposed method is an alternative for establishing secured
guarantees data transmission by acknowledgment services and communications. But for message transfer in e-commerce
when packets are lost, retransmission happens. UDP sends platforms, encryption is not stable.
packets without confirmation and multiple packets are lost in The encryption techniques in Wireless Sensor Networks
a single call. The chaining encryption algorithm for low-power emphasize the need for an increase in security levels in
wide-area networks (LPWAN) uses a key that is assigned to wireless networks and a reduction in energy consumption.[9]
each message using hash functions.[2] However, the algorithm The study increases the security in the key transmission phase
fails if there is a collision in the hash function. An efficient by using dynamic keys but fails to address the location-prone
data encryption method based on Blockchain Technology attack on the dynamic key transfer. The hacker can navigate
prevents data loss and data distortion in an E-Commerce the location or use Virtual Private Network to morph the
Platform.[3] It is a simple and effective process to encrypt E- location used. Further, in the physical layer encryption against
Commerce data. In the encryption and decryption process, the chosen plaintext attacks, the algorithm discusses the
bit error rate and the packet loss rate are low. But the algorithm ciphertexts which are dynamically generated by incorporating
uses a key-based encryption structure to ensure the accuracy random input data.[10] The algorithm requires a huge space to
of data. The key transfer might be easily compromised in this store the key.
technique and might lead to a data leak. Encryptions using hill cipher usually have the key matrix
Chinese cryptographic bases are used in an encryption generated randomly. In certain scenarios, the key matrix is
method to secure instant messaging by exploiting blockchain generated using sequential advancement and predetermined
and machine learning algorithms.[4] This involves the transfer by permuted procedures.[11] An algorithm based on a public
of both the private and public keys of the users for message key is proposed for healthcare where a receiver can use a
authentication. A cryptographic hash is used for message private key system and the data is sent to the cloud.[12]
integrity and message encryption is used for protecting the Unfortunately, the listed studies are very much prone to public
privacy of users. Since this technique involves a private key, and private key attacks.
the key must be transferred before the message and hence A concept called designated cloud server using
information can be tapped during the transfer. Cryptography authenticated encryption technique is used for the remote
and graphs are explored vastly in the encryption techniques storage system.[13] The authenticated encryption technique
and public-key encryption methods are proposed based on a proposed doesn’t allow an attacker to decipher the text. A
graph.[5] Here, the graph generated from the given message is modification of the Cayley-Purser algorithm is carried out
encrypted. The encryption is based on the properties of the using a general linear group instead of a Galois field.[14] This
matrices which are generated from the graph. The properties ensures the key of the algorithm is harder to obtain but doesn't
of the graphical codes are used for decrypting the message. eliminate the key. A compression encryption scheme that
These proposed algorithms have key generation, sender-side focuses on encrypting multimedia content uses Latin Square
graph generation, and receiver-side decryption. This might Cipher and symmetric keys.[15] Although the algorithm
prove costly as the key is used to form the matrix for the promises high speed, the Latin Square Cipher uses a short key.
required graph in the encryption process. Asymmetric image encryption which uses the Massey Omura
In a hybrid cryptosystem, which is a combination of Hill scheme requires the sender and the receiver to agree on public
Cipher and Elgamal, the encryption is carried out using Hill parameters before the encryption begins.[16] An optical
Cipher and the Hill cipher key is encrypted using Asymmetric encryption algorithm involving a focus tunable lens uses the
Elgamal.[6] The running time of this algorithm is directly fractional order of the transform as an encryption parameter.[17]
proportional to the length of the text used during the In the novel color image encryption process, the encryption

280 | Eng. Sci., 2022, 19, 279-284 © Engineered Science Publisher LLC 2022
Engineered Science Research article

and inverse processes are put forward for large key space to namely S and R respectively.
counter brute force attacks.[18] For encryptions ranging across
multi-users, a Multi Key Homomorphic encryption supports
an optimized re-linearization algorithm that uses rescaling
technology.[19] A new lightweight block cipher called
SCENERY is applied to hardware and software platforms.[20]
The encryption process of SCENERY uses an 80-bit key. All
these methods involve key exchange for the encryption-
decryption process.

2. Proposed encryption method

A good communication channel is determined by its strong
encryption algorithm. The message transmitted will be
subjected to various attacks by hackers. A small loophole in
the encryption process can lead to compromising sensitive
data. The encryption method proposed in the present study is
depicted in Fig. 1 and it works as follows:
Two polynomials—generator polynomial (G(x)) and
producer polynomial (P(x)) should be selected. The Functions
of G(x) and P(x) are carefully computed after multiple trailed
runs. A linear equation (3x-4) is taken for G(x) (Equation (1))
because the generator function runs multiple times. Taking
anything more than a binomial tends to increase the time
complexity of the algorithm. The coefficients of the binomial
are 3 and 4 as that is the maximum integer number that can be
assigned without increasing the length of the string during
encryption. For the same reason, there is a '-' (minus) sign to
bring down the length of the encrypted string. A quadratic
equation (x2-x+1) is taken for P(x) (Equation (2)) because P(x)
has lesser iterations than G(x) and can afford to stay in the
quadratic equation. The goal is to have a higher degree
equation as it brings down the possibility of brute force.
Several trial runs were conducted for P(x) and with an effort
to reduce the length of the encrypted string, the '-' (minus) is
used as mentioned before.
The iterator value x for the polynomials is computed from
the IP addresses of the sender and the receiver. The initial
values P(1) and G(1) are computed after passing the
polynomials through the Euler’s constant. The subsequent
values of G(y) and P(y) for 1 < y < = 256 are computed from
G(y-1). If P(y) > 99999, only the first 5 digits are taken to
make sure the message is not long. A cross table is constructed
mapping each of the 256 American Standard Code for
Information Interchange (ASCII) characters to one distinct P(y)
value. Hence, the message is replaced by a numerical value or
string of decimal digits after substituting the values from the
cross table for every character in the message. This resultant
encrypted message may end up being too large. So, a complex Fig. 1 Flowchart of the proposed algorithm.
cipher table is constructed, which maps a group of two or three
digits to a single ASCII character. The method is explained in Step 2: Obtain Generator G(x) and Producer P(x) using
detail in the following sub-sections. One example is also given Equations (1) and (2), respectively.
as an illustration.
𝐺(𝑥) = 3𝑥 − 4 (1)
2
2.1 Encryption 𝑃(𝑥) = 𝑥 − 𝑥 + 1 (2)
Step 1: Obtain the IP addresses of the sender and the receiver Step 3: Subtract each byte of the receiver IP address from the

© Engineered Science Publisher LLC 2022 Eng. Sci., 2022, 19, 279-284 | 281
Research article Engineered Science

corresponding byte of the sender IP address and add the and the producer polynomial. Guessing these polynomial
resultant differences to get the iterator x. functions will not work for the man-in-the-middle as the first
Step 4: Substitute each digit of the iterator x with the value iterator has to be fed to the generator polynomial, without
from the corresponding positions after the decimal point of the which the algorithm will not work. The Euler’s constant is an
Euler’s constant (2.718281828459045235360287…) essential part of the algorithm and finding out the usage of the
Step 5: Compute initial values using Equations (3) and (4): same is highly unlikely. Moreover, the producer polynomial is
𝐺(1) = 𝐺(𝑥) (3) linked to the generator polynomial and if the connection is not
laid out, decoding the message is tedious. In addition, another
𝑃(1) = P(G(1)) (4)
cryptic table is used to morph the already encrypted text, and
Step 6: Compute 256 values for P corresponding to all the 256 hence, it is a Herculean task for the hacker to obtain the
ASCII characters and store it in the cipher table with the following and do them in the right sequence. In conclusion, it
mapping P(1) for a, P(2) for b, and so on. Hence, the cipher is very hard to tap the components of the algorithm and lay the
table will have 2 columns - a list of alphabets and characters algorithm in the correct sequence to break the encryption.
in one column and its corresponding P(y) values in the other The importance of combining the IP address with the
column. P(1) is computed using Equation (1). Since there are Euler’s constant, an irrational number is explained in this
256 characters, P(2) to P(256) are computed as follows paragraph. If an irrational number like Euler’s constant,
Equations (5) and (6). For values of y = 2 to y < = 256: without the IP address, is used in the initial encryption phase,
𝐺(𝑦) = 𝐺(𝐺(𝑦 − 1)) (5) all the letters will have a different value. The drawback is that
𝑃(𝑦) = P(G(y)) (6) when the hacker taps one million messages and maps them,
If P(y) > 99999, take the first 5 digits. he/she will find a pattern with all ‘A's in the message having
Step 7: Obtain the cipher by substituting from the cipher table the same corresponding encrypted text, all ‘B's in the message
for each character in the plaintext. If the cipher length is to be having the same corresponding encrypted text, and so on.
reduced, construct the complex cipher table and obtain the When the IP addresses are included, a communication
reduced ciphertext as follows. message between Sender 1 and Receiver 1 will have a different
Step 8: If the Length (cipher) % 2 = 0, then group all the digits corresponding ciphertext when compared to a communication
in the cipher in a group of 2 digits else, group the message between Sender 2 and Receiver 2 as the IP addresses
(Length(cipher) – 1) digits in a group of 2 digits. are all different. The encryption algorithm based on a chaotic
Construct the complex cipher table by mapping each pair map has 4 values as keys and the key is randomly chosen
of digits to the printable characters. If the length of the cipher between these values.[8] This technique makes it difficult to
is odd, keep the unpaired digit as it is in the final ciphertext. decrypt but if the hacker has more data, he/she can segregate
Fig. 1 explains the processing of the IP addresses to get the and map it. In the present study, the IP addresses deal with this
iterator x and encrypt the message using the cipher table as issue and give out a different ciphertext for the same character
described in steps 1 through 8. as the IP addresses change during every communication
passage.
2.2 Decryption
Obtain the decipher table using IP addresses and the complex 3. Result in analysis
decipher table similar to the complex cipher table at the sender The algorithm is analyzed by plotting the non-linearity graph
side. If the ciphertext received contains alphabets and other of P(y) and by computing the time complexity. This algorithm
characters, use the complex decipher table to get the message was tested in a system with a Lenovo i5 processor, 256 GB
containing digits, and then group them in a sequence of 5 digits. Solid State Drive (SSD), and 8 GB Random Access Memory
Now, use the decipher table to get the plaintext. (RAM).
The algorithm is illustrated with examples along with a
cipher table and complex cipher table in the Supporting 3.1 Non-linearity
document. The plot of P(y) values for the alphabets is shown in Fig. 2
explaining the random behavior of P(y) corresponding to the
2.3 Security Analysis alphabets used. From the graph, it is evident that the P(y)
The algorithm involves the IP addresses of the sender and the values do not form linear lines and hence are not prone to
receiver. Two equations namely the generator polynomial and value-range guessing and mapping of cryptic messages. One
producer polynomial as mentioned in Equations (1) and (2) are common technique of breaking encryption algorithms is by
also used. Once the message is substituted for the generated predicting values after analyzing the patterns of the algorithm.
words from the producer polynomial, a layer of morphing is As plotted in the graph, for the algorithm in the present study,
carried out with the help of a cipher table before the message the values are completely different for every conversation
is finally transmitted. happening between two distinct people. In addition to this, the
Even if the hackers tap the IP addresses of the sender and result of the first phase of encryption goes through a complex
the receiver, it will be difficult for them to predict the generator cipher table and thus increases the security of the message.

282 | Eng. Sci., 2022, 19, 279-284 © Engineered Science Publisher LLC 2022
Engineered Science Research article

Hence, it becomes difficult for the man-in-the-middle to the message.

understand the sequence and predict the next value.
Using brute force to decode the encryption algorithm proposed 3.3 Running time of encryption
in the present study, will take multiple iterations and it has to The running time of encryption when executed in an i5
go through two table mapping. It would result in a guessing processor with 8GB RAM, is shown in Table 1 with different
game and no pattern can be drawn as the P(y) values when sample plaintexts.
plotted on a graph as in Fig. 2 gives no significant inference.
Table 1. Running Time of Encryption.
100000 Sl. No. Plaintext Running Time
90000 1. Dogs 0.002 s
80000 2. Horses 0.003 s
70000 3. Hello! How are you? 0.005 s
P(y) values

60000
50000
40000 It is observed that the running time increases as the length
30000 of the plain text increases. Similar characteristic traits of the
20000
running time are recorded in Blockchain Technology to
prevent data loss and data distortion in an E-Commerce
10000
Platform[3] and Modified Permutation Composition
0
algorithms[7] where the running time is directly proportional to
a c e g i k m o q s u w y
the length of the plaintext.
Alphabets Used Studies proposed Blockchain Technology to prevent data
loss and data distortion in an E-Commerce Platform,[3] instant
Fig. 2 The unpredictable graph obtained from values of P(y). messaging by exploiting blockchain and machine learning
algorithms,[4] and Identity-Based Encryption[12] involve either
3.2 Time Complexity a public key or a private key or both during the encryption
The algorithm is implemented and tested end to end in the C++ process. The algorithm implemented in the present study
programming language. The time complexity is computed as completely omits the concept of using keys and secures the
O(n) which is supported by the following steps: transfer of the message. The algorithm used in the three-pass
Step 1: To get the IP addresses of the sender and the receiver protocol scheme for data security cannot be signed digitally.[7]
as 4 octets each (1 octet = 8 bits), the Time Complexity = O(1). The encryption method proposed in the present study can be
Step 2: To calculate iterator x, Equations (7) to (11) are used. signed digitally using various hashing techniques if required.
𝑥1 = 𝑆𝑒𝑛𝑑𝑒𝑟𝐼𝑃 𝑂𝑐𝑡𝑒𝑡1 − 𝑅𝑒𝑐𝑒𝑖𝑣𝑒𝑟𝐼𝑃 𝑂𝑐𝑡𝑒𝑡1 (7)
𝑥2 = 𝑆𝑒𝑛𝑑𝑒𝑟𝐼𝑃 𝑂𝑐𝑡𝑒𝑡2 − 𝑅𝑒𝑐𝑒𝑖𝑣𝑒𝑟𝐼𝑃 𝑂𝑐𝑡𝑒𝑡2 (8) 4. Applications
𝑥3 = 𝑆𝑒𝑛𝑑𝑒𝑟𝐼𝑃 𝑂𝑐𝑡𝑒𝑡3 − 𝑅𝑒𝑐𝑒𝑖𝑣𝑒𝑟𝐼𝑃 𝑂𝑐𝑡𝑒𝑡3 (9) This proposed algorithm can be used for various purposes like
satellite communication or rockets in space where the
𝑥4 = 𝑆𝑒𝑛𝑑𝑒𝑟𝐼𝑃 𝑂𝑐𝑡𝑒𝑡4 − 𝑅𝑒𝑐𝑒𝑖𝑣𝑒𝑟𝐼𝑃 𝑂𝑐𝑡𝑒𝑡4 (10)
scientists are aware of the IP address of the satellite or the
𝑥 = 𝑥1 + 𝑥2 + 𝑥3 + 𝑥4; (11) rocket. They will be able to send and receive the message
Hence, Time Complexity = O(1) successfully without any other country attempting to get their
Step 3: To compute G(1) and P(1) as G(1) = G(x) and P(1) = feed data. The data is very sensitive and a man-in-the-middle
P(G(1)), the Time Complexity = O(1). attack in such cases would be expensive. To secure high-level
Step 4: To compute 256 values for P as G(y) = G(G(y-1)) and information in such scenarios, the algorithm proposed in the
P(y) = P(G(y)), the Time Complexity = O(n). current study can be used.
Step 5: To construct the complex cipher table, the Time End to End encryption has been widely used these days by
Complexity = O(1). various messaging mobile applications and hence the
Hence, the worst-case time complexity from the analysis is proposed algorithm can provide the required backend
computed as O(n). A similar time complexity of O(n) was encryption layer for a new application. It can also serve as an
found in VMPC[7] and Identity-Based Encryption with a additional layer of encryption over the already existing cryptic
Filtered Equality Test.[12] In the case of Identity-Based system of the application.
encryption techniques, the encryption process takes a longer
time than decryption but in the present study, both encryption 5. Future scope and conclusion
and decryption process complexity was found to be O(n). In One of the disadvantages of this method is that the sender and
location-based data encryption for wireless sensor networks[9] the receiver should know each other's IP addresses. The future
and chaotic encryption algorithm,[10] the time complexity was scope is to make the cipher generation more complex. Secure
found to be 4N and 2N, respectively, where N is the length of transmission of data involves advanced cryptography. To

© Engineered Science Publisher LLC 2022 Eng. Sci., 2022, 19, 279-284 | 283
Research article Engineered Science

ensure the safety of data, a huge leap towards encryption of Computing and Communication Technologies (WCCCT), 2017, 1,
data is essential. Key-based encryption techniques become 51-54, doi: 10.1109/WCCCT.2016.22.
very vulnerable to man-in-the-middle attacks. Furthermore, [12] Y. Ming, E. Wang, Sensors, 2019, 19, 1-22, doi:
with the advancement in computing, brute force can decrypt 10.3390/s19143046.
the message once the key is obtained. The algorithm is [13] V. Chenam, S. Ali, Computer Standards & Interfaces, 2022,
proposed to counter the vulnerabilities posed by key transfer 81, 1-21, doi: 10.1016/j.csi.2021.103603.
and hence an effort is taken to abolish the transfer of key to [14] S. Khlebus, Faris Hasoun, K. Rajaa, B. Sabri, International
make it less vulnerable. In the proposed algorithm, the Journal of Nonlinear Analysis and Applications, 2022, 13, 707-
message is encrypted by constructing a generator polynomial 716, doi: 10.22075/ijnaa.2022.5559.
and producer polynomial without using a key. Corresponding [15] A. Bensaoud, J. Kalita, Journal of Information Security and
new characters are replaced in the message and the obtained Applications, 2022, 64, 1-21, doi: 10.1016/j.jisa.2021.10303.
ciphertext undergoes a reduction in length when values from [16] K. G. Abdulhussein, N. M. Yasin, I. J. Hasan, International
the complex cipher table are substituted. The decryption Journal of Electrical and Computer Engineering, 2022, 12,
process involves similar steps and thus the message transfer is 103057, doi: 10.1016/j.jisa.2021.103057.
not compromised. The complexity of the algorithm is O(n). [17] J. Alexis Jaramillo-Osorio, W. Torres-Sepúlveda, A. Velez-
The graph generated for the Producer polynomial (Fig. 2) Zea, A. Mira-Agudelo, J. Fredy Barrera-Ramírez, R. Torroba,
shows the random behavior of the encryption process. Hence, Optics & Laser Technology, 2022, 148, 107689, doi:
the message is tightly secured while being transmitted in the 10.1016/j.optlastec.2021.107689.
communication passage. [18] C. F. Duan, J. Zhou, L. H. Gong, J. Y. Wu, N. R. Zhou, Optics
and Lasers in Engineering, 2022, 150, 106881, doi:
Conflict of Interest 10.1016/j.optlaseng.2021.106881.
The authors declare no conflict of interest. [19] X. Yang, S. Zheng, T. Zhou, Y. Liu, X. Che, Tsinghua Science
and Technology., 2021, 27, 642-652, doi:
Supporting information 10.26599/TST.2021.9010047.
Not applicable. [20] J. Feng, L. Li, Frontiers of Computer Science, 2021,16,
163813, doi: 10.1007/s11704-020-0115-9.
Reference
[1] S. Suherman, S. Panjaitan, A. Ginting, Journal of Physics: Author Information
Conference Series, 2019, 1235, 012032, doi: 10.1088/1742-
6596/1235/1/012032. Aakash Bharadawaj Srinivasan has a
[2] A. Bidgoly, H. Bidgoly, IEEE Sensors Journal, 2019, 19, Bachelor’s in Information Technology and is
7027-7034, doi: 10.1109/jsen.2019.2910850. currently pursuing his Masters’ in Business
[3] F. Gao, Discrete and Continuous Dynamical Systems, 2019, Administration. He has published multiple
12, 1457-1470, doi: 10.3934/dcdss.2019100. papers across cryptography, Internet of Things
[4] H. Yi, Safety Science, 2019, 120, 6–13, doi: and Natural Language Processing.
10.1016/j.ssci.2019.06.025.
[5] D. Sensarma, S. Sarma, International Journal of Innovation Hemalatha S is an Associate Professor
and Technology, 2019, 8, 2273-2279, doi: (Senior-Scale) in the Department of Computer
10.35940/ijitee.J1133.0881019. Science and Engineering, Manipal Institute of
[6] D. Rachmawati, A. Sharif, Ericko, Journal of Physics: Technology, Manipal Academy of Higher
Conference Series, 2018, 1235, 1-7, doi: 10.1088/1742- Education, Manipal, India. She received her
6596/1235/1/012074. MTech and Ph.D degree from MIT Manipal, India. Her
[7] M. Budiman, D. Rachmawati, R. Badegeil, Journal of Physics: research interests include information security, image
Conference Series, 2019, 1235, 012085, doi: 10.1088/1742- processing, computer architecture and machine learning.
6596/1235/1/012085.
[8] P. Cheng, H. Yang, P. Wei, W. Zhang, Nonlinear Dynamics, Ramathmika has Bachelor’s in Computer
2015, 79, 2121-2131, doi: 10.1007/s11071-014-1798-y. Science and Engineering from Manipal Institute
[9] H. Lin, Wireless Networks, 2015, 21, 2649-2656, doi: of Technology, Manipal Academy of Higher
10.1007/s11276-015-0938-8. Education, Manipal, India and is currently
[10] X. Yang, Z. Shen, X. Hu, W. Hu, Chaotic encryption working for an IT company.
algorithm against chosen-plaintext attacks in optical OFDM
transmission, IEEE Photonics Technology Letters, 2016, 28, Publisher’s Note: Engineered Science Publisher remains
2499-2502, doi: 10.1109/lpt.2016.2601659. neutral with regard to jurisdictional claims in published maps
[11] R. Mahendran, K. Mani, Generation of key matrix for hill and institutional affiliations.
cipher encryption using classical cipher, 2017 World Congress on

284 | Eng. Sci., 2022, 19, 279-284 © Engineered Science Publisher LLC 2022