2020 7th International Conference on Signal Processing and Integrated Networks (SPIN)

Analysis and Optimization of Advanced

Encryption Standard for the Internet of Things
Ishfaq Sultan, Bisma Javid Mir, and M. Tariq Banday*, Senior Member, IEEE
Department of Electronics and Instrumentation Technology
University of Kashmir, Srinagar, India
*
[email protected]

Abstract—The Internet of Things creates advanced application modules. The best recommended solution for the security in
systems requiring minimum human interaction by integrating IoT is lightweight cryptography which reduces the total
entities having unique addressing mechanisms with digital implementation cost related to many parameters, such as key
processing platforms by communicating in a network. The size, power consumption, area, cycle rate and throughput
Internet of Things devices being resource constrained are prone
to attacks and security breaches. Conventional cryptographic
rate. The techniques used in lightweight cryptography
procedures are not viable for Internet of Things and embedded consume small memory and requires less computation, thus
applications as the computations involved with them require using least amount of necessary resources of the nodes.
higher processing power, memory and energy. Lightweight The lightweight cryptography is broadly classified into
cryptographic techniques can be used in resource constrained
Internet of Things devices to mitigate the security issues. These two categories [12], symmetric lightweight ciphers and
lightweight cryptographic solutions use smaller key sizes, asymmetric lightweight ciphers. Symmetric lightweight
smaller block sizes and lesser number of rounds. This paper uses ciphers are faster and consume less power but are less secure
a lightweight cryptographic algorithm, AES (Advanced compared to asymmetric lightweight ciphers. Asymmetric
Encryption Algorithm) which is one of the efficient and robust ciphers on the other hand are slower and consume more
algorithms, to secure communication in the wireless sensor power. Symmetric lightweight techniques comprise of
networks. The experimentation uses UDP protocol for lightweight hash algorithms and lightweight block ciphers.
transmissions and Routing Protocol for Low-Power and Lossy The two NIST-accepted lightweight block ciphers [7] include
networks (RPL) for packet delivery from source to destination. Advanced Encryption Standard (AES) and Triple Data
Later, the performance of different variants of AES is checked
by performing the power analysis. The results demonstrate that Encryption Algorithm (TDEA) [13]. The Advanced
AES-192 and AES-256 with 8 rounds show optimized results in Encryption Algorithm has three variants that comprise of
terms of power, thus making it suitable for power constrained AES-256, AES-192, and AES-128 with key sizes 256-, 192-
devices. and 128-bit, respectively. The block size of all the variants is
same (128-bits).
Keywords—Lightweight Cryptography; UDP; User Datagram
In this paper, different modes of AES i.e., AES-ECB,
Protocol; RPL; Routing Protocol for Low-Power and Lossy
Networks; Contiki; Cooja.
AES-CBC, and AES-CTR with variable key lengths have
been implemented in Contiki OS using Cooja simulator.
I. INTRODUCTION These modes have also been tested and implemented using 8
rounds instead of 12 and 14 in 192- and 256-bit key lengths
The concept of the Internet of Things was introduced in respectively. The motes used in the simulation are Sky and Z1
1998 by Kevin Ashton [1]. Internet of Things (IoT) is an motes. Sky mote is based on Texas Instruments MSP430
interconnection of devices which have the capability of F1611 microcontroller whereas Z1 mote is based on Texas
sensing, actuating, communicating, etc. through the Internet Instruments MSP430 F2617 microcontroller. Both the motes
where human involvement is minimum. IoT is a term that is are used in low power IoT and wireless sensor networks and
allied with the notion of “future Internet” [2]. are based on IEEE 802.15.4 protocol for wireless
The IoT architecture comprises of three basic layers: communication.
physical layer which consists of sensors and actuators,
II. RELATED WORK
network layer which have communication capabilities and
application layer at the user end [3][4]. IoT is presently an Margi et al. [14] investigated the influence of operating
emergent technology that will cover every aspect of future to systems (TinyOS and Contiki) on the security of IoT and
create a much easy life for us with smart devices all around. embedded applications on the TelosB platform. Hyncica [15]
Few instances of the IoT systems are smart buildings and the authors assessed the performance of 15 block ciphers on
homes, smarter vehicles, e-health aids etc. [5]. These IoT 8-, 16- and 32-bit microcontrollers. They adopted TomCrypt
systems are commonly used today and it is expected that the LTC library which is generally suited for 32-bit platforms and
total number of nodes in 2020 will cross 212 billion [3]. is not optimized for constrained IoT devices. They
implemented the Electronic Codebook (ECB) mode of AES
Besides the heterogeneous and ubiquitous nature of
operation, which targets only confidentiality and permits a
IoT, huge number of nodes monitoring the environment
clear distinction in the performance of the plain block ciphers
makes security a key issue in IoT devices. Severe security
compared to other modes of operation. ECB mode is
challenges are growing in terms of data confidentiality,
considered insecure in the modern world because the
integrity, and authenticity. Conventional cryptographic
applications demand more advanced security measures
algorithms are slow in speed, use large key size and consume
including authentication and data integrity. Therefore, it is
huge power. Therefore, these techniques cannot be used in
necessary to evaluate the performance assessment over
resource constrained IoT applications [6]. Since the
resource constrained devices of other AES modes of
execution time of traditional cryptosystems is enormous,
operation. The authors did not address the power
several researchers have tried to decrease this time
measurements which is an important aspect as far as
[7][8][9][10][11]. But these attempts resulted in another
constrained IoT devices are concerned.
constraint i.e., cost. The hardware requirements for such a
system are very costly because of the suggested integrated
978-1-7281-5475-6/20/$31.00 ©2020 IEEE 571
Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY KURUKSHETRA. Downloaded on January 27,2025 at 06:46:04 UTC from IEEE Xplore. Restrictions apply.
 2020 7th International Conference on Signal Processing and Integrated Networks (SPIN)

Law, Doumen and Hartel [16] implemented 8 different The output state array produced by the last round is
block ciphers in 4 encryption modes on the MSP430F149 rearranged into an output block of size 128-bit.
platform. The platform is based on Texas Instruments 16-bit TABLE 1. RELATIONSHIP BETWEEN KEY SIZE AND NUMBER OF ROUNDS
RISC microcontroller. The authors focused on encryption
modes and did not investigate other conventional security Length Size Size of Total
Key
Algorithm of Key in of Block in Number
principles required by almost all IoT devices. Geovandro, Size
words Block words of rounds
Pereira and Renan [17] presented energy and time AES-256 256 8 128 4 14
benchmarks of encryption implementations for different AES-192 192 6 128 4 12
platforms and operating systems. The authors used the Intel AES-128 128 4 128 4 10
Edison platform for experimentation and gave benchmark
The block ciphers are structures used for the process of
results of symmetric cryptography, and also described the
encipherment and decipherment where a block of input data
methodology for energy consumption measurement on the
is treated as one block and produces a block of output data
said platform. Peyrard et al [18] improved the security level
that is of the same size [24]. In AES, block size is 128-bits or
in Contiki OS by demonstrating the usage of formal
16 bytes. The block cipher algorithms should have the
verification. The authors presented a case study on analytical
capability of enciphering an input data with different size
verification of encryption and decryption modules of Contiki
than the defined block size of the algorithm. It can be
OS on AES-CCM* (CCM-star), which is a variant of AES in
provided by using different modes of operation. Some of
counter with CBC-MAC mode by employing software
these modes convert the block cipher into stream cipher thus
analysis platform: Frama-C for C code. Abdelmoghni [19]
making the algorithm strong. The different modes of
investigated the energy consumption and duration of the AES
operation standardized by NIST in 2001 are Cipher FeedBack
both in software and hardware with varying key and buffer
(CFB), Cipher Block Chaining (CBC), Electronic Code Book
size on two resources constrained IoT platforms.
(ECB), Counter mode (CTR) and Output Feed Back (OFB).
6LoWPAN adaptation layer is one of the prevalent All these modes have been approved for AES [25].
solutions for adopting transmitting messages over IEEE
802.15.4 networks. Rantos, et al [20] focused on the security A. ECB (Electronic Code Book) Mode
delivered to LLN (low-power and lossy network) nodes by
exploiting 6LoWPAN adaptation setup. The authors The simplest mode of operation is Electronic Code
suggested a compression format for IPsec, which is able to Book (ECB) Mode. In this mode, the ciphertext is generated
offer end-to-end security. The format utilized AES-CCM* by dividing the plaintext in blocks and then encrypting each
(CCM-Star), while contemplating the boundaries of the block separately using the same key. The resultant ciphertext
fundamental IEEE 802.15.4 protocol. The structure proposed is also generated in blocks. If the input message is not a
by the authors offered authentication, confidentiality, and multiple of block size, it is padded such that it becomes a
integrity of data and featured low packet overhead compared multiple of block size. One of the advantages of this mode is
to similar approaches. Smeets et al. [21] proposed a that error is not propagated from one block to another
symmetric key management structure for low power wireless [26][23][27]. But this mode has a weakness that similar
sensor networks that have the capability to use damage proof plaintext blocks result in similar ciphertext blocks, thus
hardware for key distribution and generation. The authors making it unsuitable for encryption of data greater than one
pointed out that the structure made use of a trusted central block [28][26].
entity for key negotiation and no deployment knowledge was
required before enrolling to offer end-to-end security. The B. CBC (Cipher Block Chaining) Mode
assessment and implementation were performed on low To overcome the disadvantage of ECB Mode, this mode
power Zolertia Z1 hardware platform running on the Contiki implements two new ideas. The first idea is that the
operating system. encryption of all the blocks is reliant on each other in such a
III. ADVANCED ENCRYPTION STANDARD way that the ciphertext depends not only on the current block
but also on all the previous plaintext blocks. The second idea
National Institute of Standards and Technology (NIST) is that an initialization vector (IV) is used to create
in 2001 confirmed the block cipher Rijndael as the new randomness in the ciphertext [23]. The size of the IV is equal
Advanced Encryption Standard (AES) and declared it as an to the block size. In this mode, the first plaintext block is x-
absolute standard in FIPS Publication 197 [22]. Rijndael was ored with the IV and the output generated is encrypted using
given by two new Belgian scientists, Joan Daemen and the key resulting in the first ciphertext. This ciphertext is then
Vincent Rijmen. AES is the most extensively used symmetric x-ored with the next plaintext message and the result is again
cipher which means that a single key is used for the operation encrypted using the same key, thus generating the second
of encryption and decryption. It is safe against all the known ciphertext. In the similar manner, the process continues.
attacks. The key size in AES can be 128-, 192-, 256-bits and Thus, in this mode different output is generated for two
the block size is 128-bits. The number of rounds in AES with similar blocks. Also, a different result can be obtained after
key size 128-, 192-, 256-bits are 10, 12 and 14 respectively encrypting two similar messages if the IV used is different.
as shown in Table 1 [23]. The disadvantage in this mode is that an error in one block
will propagate to all other blocks which will be reflected in
The operations of AES are performed on bytes unlike
the process of decryption.
DES where operations are carried out in bits. 128-bits of
plain-text block is treated as 16 bytes in AES. These 16 bytes
are arranged as a matrix with four rows and four columns. In C. CTR (Counter) Mode
these 16 bytes, first 4 bytes occupy the first column in the In this mode, the block cipher works like a stream
matrix of size 4x4. The next 4 bytes occupy the second cipher. An input block called counter (just like the
column, the following 4 bytes occupy the third column and initialization vector) is used. The size of the counter is equal
the last 4 bytes occupy the fourth column. This 4x4 matrix is to the block size. The counter is encrypted using the key and
called the state array. In each round of AES, processing is the output is x-ored with the first plaintext block to generate
done on input state array to generate the output state array. the first ciphertext block. For the next block, the counter is

572
Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY KURUKSHETRA. Downloaded on January 27,2025 at 06:46:04 UTC from IEEE Xplore. Restrictions apply.
 2020 7th International Conference on Signal Processing and Integrated Networks (SPIN)

incremented and encrypted using the same key and then again CPU i.e., CPU power consumption for this particular cycle;
x-ored with the second plaintext block. For the following lpm i.e., LPM power consumption for this particular cycle;
blocks the counter is incremented again and again and, in this transmit i.e., transmission power consumption for this
way, the encryption of the data takes place. Just like ECB, particular cycle; listen i.e., listen power consumption for this
CTR does not propagate error from one block to another particular cycle; idle_transmit i.e., idle transmission power
[23][29]. As there is no feedback in this mode, thus an consumption for this particular cycle; idle_listen i.e., idle
efficient implementation of this mode is possible in both listen power consumption for this particular cycle;
software and hardware by exploiting the use of parallelism
[28]. This implementation was done in Cooja using Sky and
Z1 motes. Tmote Sky is an ultra-low power wireless module
IV. IMPLEMENTATION based on Texas Instruments MSP430 F1611 microcontroller
extensively used in wireless sensor networks and rapid
The work carried out in this paper has been done in an
application prototyping. It consists of 2.4GHz IEEE 802.15.4
open source operating system called Contiki. This operating
wireless transceiver having interoperability with other IEEE
system is best suitable for embedded devices with less
memory and is also used in wireless sensor networks. Contiki 802.15.4 devices. Z1 mote is a low power wireless hardware
operating system is extremely portable, supports multitasking platform based on Texas MSP430 F2617 microcontroller
and is best suited for microcontrollers with small memory. It complaint with IEEE 802.15.4 and Zigbee protocols used
has 40 KBs of ROM and 2KBs of RAM in its typical widely in wireless sensor networks and low power IoT
configuration. The code in Contiki consumes memory of the applications. In this paper, the power consumption of
order of kilobytes and it can be configured to be as low as different AES encryption algorithms is analyzed using Sky
tens of bytes. Both IPv6 and IPv4 are supported in Contiki. and Z1 motes in Cooja simulator using the Contiki operating
Contiki-OS presented the concept of using Internet Protocol system.
communication in low-power sensor networks. Rime and uIP The implementation uses a simple example of RPL-UDP
are the two communication stacks in Contiki. Rime has been present in Contiki operating system where transmission is
aimed for low-power radios as it is a lightweight
done using UDP protocol and packet delivery from source to
communication stack. It offers basic communication that
destination is done by RPL protocol. The implementation is
ranges from local area broadcast to reliable multi-hop bulk
data flooding. Similarly, uIP being a small TCP/IP stack is done in Cooja simulator using Sky and Z1 motes. User
responsible for communication over the Internet in Contiki. Datagram Protocol, which is part of Internet Protocol suite
Contiki-OS is supported by a wide range of platforms that (UDP/IP suite) is unreliable and connectionless protocol and
include embedded microcontrollers like ARM, AVR, therefore does not require a connection prior to data transfer.
msp430, pic32, etc., old home computers, and many more. The example in our case simulates a client and a server node.
The client node routes data and sends it to the server node
The simulator for Contiki OS is called Cooja. The using the RPL routing protocol. While sending the data from
motes in Contiki are simulated in Cooja which is a real client, the data is sent to buffer first where the encryption
compiled and executing Contiki system. Cooja works by operation takes place. The buffer is then transmitted to the
compiling Contiki for the native platform as a shared library
server using the UDP protocol. The encrypted message in the
and using JNI (Java Native Interface) loads the library into
Java. Different kinds of sensor nodes can be a part of Cooja buffer is decrypted using the decryption algorithm at the
simulation wherein different nodes can incorporate different server end after it is received. The modification of the data in
Contiki libraries. Few functions help Cooja to analyse and the buffer after encryption or decryption is done using the
control Contiki OS. A tool called powertrace is used in different modes of the AES algorithm. The same message of
Contiki application to measure the amount of power length 48 bytes is encrypted and transmitted using different
consumed. To use this tool in any Contiki application, simply key sizes and encryption modes including AES-ECB, AES-
edit the Makefile of the application using: CBC, and AES-CTR. The message is also encrypted with a
modified 192- and 256-bit key using only 8 rounds. The
APPS += powertrace transmission power and total power consumption in each of
Another change that needs to be done, is in the main the above encryption modes are computed and compared.
program of the application:
V. RESULTS AND DISCUSSION
This library is added in the beginning of the code:
The experimentation shows the analysis and
#include "powertrace.h"
optimization of the AES algorithm on the basis of power
This function is added in the PROCESS_THREAD after consumption which is done using the powertrace application.
the command PROCESS_BEGIN: The powertrace application is embedded in the Contiki
powertrace_start (CLOCK_SECOND * 10); operating system which generates different power parameters
like CPU, LPM, transmission, reception, idle transmission,
Where 10 is the runtime. It implies that the values of idle reception, etc. The simulation has been carried out in
different parameters of power consumption will be printed Cooja simulator where the different AES types with
after 10 seconds repeatedly [13]. appropriate rounds and key sizes have been analysed. The
power analysis for all types of AES has been done on the
The printed parameters include clock_time i.e., clock
payload of 48 bytes transmitted using the RPL/UDP protocol.
time; rimeaddr i.e., rime address; seqno i.e., sequence
The five configurations include AES with 128-, 192-, 256-bit
number; all_cpu i.e., collected CPU power consumption in
keys, and AES with 192- and 256-bit keys using only 8
active mode; all_lpm i.e., collected power consumption in
rounds instead of 12 and 14 respectively. The experimental
Low Power Mode (LPM); all_transmit i.e., collected power
implementation is done by using Sky and Z1 motes as the
consumption while transmission; all_listen i.e., collected
target devices in the simulation environment. The input
power consumption while reception; all_idle_transmit i.e.,
voltage of the Sky mote is taken as 3V, and the transmission
collected idle transmission power consumption;
and reception peak currents are taken as 21mA and 23mA
all_idle_listen i.e., collected idle listen power consumption;
respectively. The active mode and the low power mode
573
Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY KURUKSHETRA. Downloaded on January 27,2025 at 06:46:04 UTC from IEEE Xplore. Restrictions apply.
 2020 7th International Conference on Signal Processing and Integrated Networks (SPIN)

currents are taken as 600μA and 3μA respectively. Similarly, Key Power consumption Power consumption (mW)
for Z1 mote the input voltage is taken as 3V, and the Length (mW) for different modes for different modes of
(bits) of AES in Sky mote AES in Z1 mote
transmission and reception peak currents are taken as 17.4mA ECB CBC CTR ECB CBC CTR
and 18.8mA respectively. The active mode and the low power 128 0.618 0.629 0.645 0.317 0.358 0.410
mode currents are taken as 0.5mA and 0.5μA respectively. 192 0.688 0.694 0.712 0.454 0.484 0.512
256 0.695 0.710 0.744 0.989 1.241 1.861
192(8R) 0.658 0.669 0.687 0.331 0.364 0.412
256(8R) 0.675 0.684 0.710 0.584 0.696 0.981

CONCLUSION

Lightweight block ciphers play a key role as far as

security of resource-constrained IoT devices is concerned. In
this paper, we have implemented AES block cipher in a
simulation environment using Cooja simulator of the Contiki
operating system. Three different modes of AES are
implemented which include AES-ECB, AES-CBC, and AES
CTR on Zolertia Z1 and Sky motes in Cooja. The power
consumption is calculated for five variable key
configurations on the said modes of AES operation. The
results show that AES-ECB is best-suited mode as far as
power consumption is concerned followed by AES-CBC and
FIGURE 1: TOTAL POWER CONSUMPTION OF DIFFERENT ENCRYPTION TYPES
USING VARIABLE KEY LENGTHS IN COOJA SIMULATOR USING SKY MOTE.
AES-CTR modes. The modified key lengths, i.e. 192- and
256-bit with 8 rounds show a reduction in power
Figure 1 shows the power consumption (in milliwatts) consumption compared to their counterparts. Hence, these
for Sky mote in different AES encryption types using five key modifier keys can be used in resource-constrained IoT
configurations. As evident from the figure AES-ECB devices and can also be helpful against brute-force and
consumes least power followed by AES-CBC and AES-CTR differential cryptanalytic attacks because of the larger key
modes. The power consumption is least at 128-bit key length size.
in all the three modes. The modified key versions of 192- and
256-bit with only 8 rounds consume less power compared to ACKNOWLEDGEMENT
their standard versions using 12 and 14 rounds respectively.
The power consumption in modified AES-256 is comparable This work has been supported by the Ministry of
to standard 128-bit key length in all three encryption types. Electronics and Information Technology, Govt. of India
under its project grant no.12 (2)/2017-CSRD.

REFERENCES

[1] Ashton, K. (2009). That ‘internet of things’ thing, RFID Journal, Vol.
22, no. 7, pp. 97-114.
[2] Bhumi, N., Tushar, C. (2015). Study of various Internet of Things
platforms, International Journal of Computer Science & Engineering
Survey (IJCSES) Vol.6, No.6, December 2015.
[3] Al-Fuqaha, A., Guizani, M., Mohammadi, M., Aledhari, M., Ayyash,
M. (2015). Internet of Things: A Survey on Enabling Technologies,
Protocols, and Applications, IEEE Communications Surveys &
Tutorials, vol. 17, no. 4, pp. 2347-2376.
[4] Sethi P., Sarangi, S. R. (2017). Internet of Things: Architectures,
Protocols, and Applications, Hindawi Journal of Electrical and
Computer Engineering, Article ID 9324035, 25 pages, 2017.
[5] Arsalan, M. N., Niraj, K. J. (2017). A Comprehensive Study of
FIGURE 2: TOTAL POWER CONSUMPTION OF DIFFERENT ENCRYPTION TYPES
Security of Internet-of-Things, IEEE Transactions on emerging topics
USING VARIABLE KEY LENGTHS IN COOJA SIMULATOR USING Z1 MOTE.
in computing, vol. 5, no. 4, pp. 586-602, 1 Oct. – Dec. 2017.
Figure 2 shows the power consumption (in milliwatts) [6] Tapalina B. (2013). LICRYPT: Lightweight Cryptography
Technique for Securing Smart Objects in Internet of Things
for Z1 mote in different AES encryption types using five key Environment, CSI Communications, May 2013.
configurations. AES-ECB consumes the least power [7] McKay, K. A., Bassham, L., Turan, M. S., Mouha, N. (2016). Report
followed by AES-CBC and AES-CTR modes. The power on lightweight cryptography. NIST DRAFT, NISTIR, pp 1–29.
consumption is least at 128-bit key length in all the three [8] Elbirt, A. J. (2007). Fast and efficient implementation of AES via
modes. As shown in Figure 2 last two key versions of 192- instruction set extensions, Proc. of 21st International Conference on
and 256-bit with only 8 rounds consume less power compared Advanced Information Networking and Applications Workshops
AINAW'07, vol. 1, pp. 396-403.
to their normal versions using 12 and 14 rounds respectively. [9] Hodjat, Alireza, and Verbauwhede, I. (2004). Interfacing a high-speed
The power consumption in modified AES-256 is comparable crypto accelerator to an embedded CPU. In Signals, Systems, and
to standard 128-bit key length in all three encryption types. Computers, Proc. of Thirty-Eighth IEEE Asilomar Conference, vol. 1,
Table II shows the comparison of power consumption values pp. 488-492.
in three encryption types using different key configurations [10] Grabher, P., Großschädl, J., Page, D. (2008). Light-weight instruction
set extensions for bit-sliced cryptography, Cryptographic Hardware
in Cooja between Sky and Z1 motes. As is evident from the and Embedded Systems–CHES, pp. 331-345.
table Z1 Mote is more efficient than Sky Mote for all the [11] O’Melia, Sean, Elbirt, A. J. (2008). Instruction Set Extensions for
encryption modes. Enhancing the Performance of Symmetric-Key Cryptography,
Proceedings of Computer Security Applications Conference, pp. 465-
TABLE 2. RELATIONSHIP BETWEEN KEY SIZE AND NUMBER OF ROUNDS 474.

574
Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY KURUKSHETRA. Downloaded on January 27,2025 at 06:46:04 UTC from IEEE Xplore. Restrictions apply.
 2020 7th International Conference on Signal Processing and Integrated Networks (SPIN)
[12] Hammad, B. T., Jamil, N., Rusli, M. E., Reza Z’aba, M., Ahmed, I.
T. (2007). Implementation of Lightweight Cryptographic Primitives,
Journal of Theoretical and Applied Information Technology, vol. 95,
no. 20.
[13] Barker, W. C., Barker, E. (2012). Recommendation for the Triple
Data Encryption Algorithm (TDEA) Block Cipher, NIST Special
Publication (SP) 800-67.
[14] Margi, C. B., De Oliveira, B. T., De Sousa, G. T., et al., (2010). Impact
of operating systems on Wireless Sensor Networks (security)
applications and testbeds, Proceedings of the 2010 19th International
Conference on Computer Communications and Networks, (ICCCN
’10), August 2010.
[15] Hyncica, O., Kucera, P., Honzik, P., Fiedler, P. (2011). Performance
evaluation of symmetric cryptography in embedded systems,
Proceedings of the 6th IEEE International Conference on Intelligent
Data Acquisition and Advanced Computing Systems: Technology and
Applications, IDAACS’2011, pp. 277–282, September 2011.
[16] Law, Y.W., Doumen, J., Hartel, P. (2006). Survey and benchmark of
block ciphers for wireless sensor networks, ACM Transactions on
Sensor Networks, vol. 2, no. 1, pp. 65–93.
[17] Geovandro, C. C. F., Pereira, Renan, C. A., Alves, Felipe L. da Silva,
Roberto M., Azevedo, Bruno C., Albertini, Cíntia B. Margi. (2017).
Performance Evaluation of Cryptographic Algorithms over IoT
Platforms and Operating Systems, Security and Communication
Networks, vol. 2017, Article ID 2046735, 16 pages.
[18] Peyrard, A., Kosmatov, N., Duquennoy, S., Lille, I., Raza, S. (2018).
Towards Formal Verification of Contiki: Analysis of the AES-CCM*
Modules with Frama-C. In Proceedings of the 2018 International
Conference on Embedded Wireless Systems and Networks (EWSN
’18). Junction Publishing, USA, 264-269.
[19] Abdelmoghni, T., Mohamed, O. Z., Billel, B., Mohamed, M.,
Sidahmed, L. (2018). Implementation of AES Coprocessor for
Wireless Sensor Networks, 2018 International Conference on Applied
Smart Systems (ICASS), Medea, Algeria, 2018, pp. 1-5.
[20] Rantos, K., Papanikolaou, A., Manifavas, C. (2013). IPsec over IEEE
802.15.4 for low power and lossy network, proceedings of the 11th
ACM International Symposium on Mobility Management and
Wireless Access (MobiWac '13). ACM, New York, NY, USA, 59-64.
[21] Smeets, Ruben, Aerts, Kris, Mentens, Nele, Braeken, An, Segers,
Laurent, and Touhafi, Abdellah. (2014). Cryptographic Key
Management Architecture for Dynamic 6LowPan Networks,
proceedings of the 9th International Conference on Applied
Informatics 1 (2014): 219-29.
[22] FIPS. (2002). Advanced Encryption Standard, Federal Information
Processing Standard (FIPS) Publication 198, 2002.
[23] Paar, C., Pelzl, J. (2010). Understanding Cryptography: Textbook for
Students and Practitioners, London, Springer, 2010.
[24] Blazhevski, D., Bozhinovski, A., Stojchevska, B., Pachovski, V.
(2013). Modes of Operation of The AES Algorithm, Proc. of 10th
Conference for Informatics and Information Technology (CIIT 2013).
[25] Dworkin, N. (2001). Recommendation for Block Cipher Modes of
Operation, Methods, and Techniques, NIST Special Publication 800-
38A Edition 2001.
[26] Stallings, W. (2005). Cryptography and Network Security: Principles
and Practices, Fourth Edition. NJ, Prentice-Hall, 2005.
[27] Rogaway, P. (2011). Evaluation of Some Blockcipher Modes of
Operation. Cryptography Research and Evaluation Committees
(CRYPTREC), 2011.
[28] Douglas, R., S. (2006). Cryptography Theory and Practice, Chapman
and Hall/CRC. 2006.
[29] Henriquez, F. R., Saqib, N. A., Perez, A. D., Koc, C. K. (2006).
Cryptographic Algorithms on Reconfigurable Hardware, Springer,
2006.

575
Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY KURUKSHETRA. Downloaded on January 27,2025 at 06:46:04 UTC from IEEE Xplore. Restrictions apply.