Energy Requirements in Cryptographic Mechanisms for Secure Wireless Sensor Networks: An Overview

Full text

Energy Requirements in Cryptographic Mechanisms for Secure Wireless Sensor
Networks: An Overview
Mompoloki Pule, Rodrigo Jamisola, Frank Ibikunle
Botswana International University of Science and Technology
Electrical, Electronics and Telecomms Engineering Department
Palapye, Botswana
ABSTRACT
Wireless Sensor Networks (WSNs) have gained
popularity in recent years. This is because they have great
potential to provide a promising infrastructure for
numerous applications. The rapid deployment and
reduction in cost of broadband internet connectivity has
made it affordable to have these networks exchange and
manage information over the public network. To use
conventional security architectures in this regard pose a
major challenge since available cryptographic algorithms
are computationally intensive. On the other hand WSN
nodes are resource constrained in terms of computational
power, storage memory, communication bandwidth, and
battery power/energy. However the energy constraint of
all is very crucial and needs to be addressed since WSN
nodes are typically power limited. The performance of
WSNs can be improved by introducing powerful
processors with large memory capacities and high
bandwidth radio technologies demanding additional
energy requirements. It is well known that communication
overheads consume more energy than performing
cryptographic computations. Thus additional control
overheads introduced on top of the data plane by
cryptographic mechanisms come at a huge cost. This
paper provides an overview of existing cryptographic
mechanisms applicable to WSNs along with their energy
requirements, strengths and weaknesses.
KEY WORDS
Wireless Sensor Networks, Cryptographic Algorithms,
Energy Requirements
1. Introduction
WSNs have always been designed to implement only the
requirements of a dedicated function, which is why they
have always retained their traditional small form factor
and have always had limited resources in terms of
computational power, storage memory, communication
bandwidth and battery power. It is because of these
features that their costs have been greatly reduced making
them effectively inexpensive.
An attempt to enhance their resources by employing more
powerful processors, large memory capacities and high
bandwidth radio technologies effectively result in bulky
sensor nodes with increased power requirements. This in
turn defeats the purpose for which these devices were
initially designed because it introduces a considerable
investment for a device which should relatively be of low
cost.
Broadband internet connectivity has rapidly become
cheap and ubiquitous, and as a result it has become very
affordable for a lot of electronic devices to use the public
network (internet) to send their information. The number
of devices connected to the internet exceeded the number
of people on earth in 2008/2009, while in 2010 the ratio
of connected devices per person was 1.84:1 [1], [2]. Due
to this exponential growth, CISCO now estimates that by
2020 at least 20 billion devices will be connected to the
internet [1]–[3].
This new technological paradigm is referred to as the
“Internet of Things” (IoT). Authors in [1], [2] describe the
IoT as a system where items in our physical world are
equipped with sensors that allow them to connect to the
internet through wired or wireless means. Such an
implementation results in a global network of smart
objects equipped with embedded electronics, software and
connectivity which enables them to exchange data
through the public network.
As we connect more of these devices to the internet, it is
very important to simultaneously implement reliable
security architectures.
WSNs play a major role in this new technological
revolution due to their numerous applications [4], [5].

These networks are made up of two main components
namely node and base station/sink. The node is an
autonomous device normally equipped with sensors that
perform a collaborative measurement process. The base
station captures and processes all the data from the nodes
and sometimes provides gateway services to communicate
with the public network [6].
Figure 1 shows the IoT enabled WSN architecture. The
key component is the WSN node. It is equipped with
sensing, processing and communication capabilities to
monitor the parameters of the intended application. Figure
2 shows a typical WSN node architecture. Nodes are
typically powered with batteries hence making energy
consumption an issue to take into account when
implementing such networks [4].
Figure 1: Wireless Sensor Network Architecture [7]
Figure 2: WSN Node [7]
2. WSN Constraints and Security Issues
WSNs face a lot of security challenges due to the nature
of their deployment. They are normally distributed and
deployed in remote areas where they are left unattended,
making them vulnerable to physical attacks such as node
capture and tampering [4], [8]. The implementation of
reliable security mechanisms to counteract such attacks is
an aspect of prime significance.
Since the inception of the IoT, it has become affordable
for WSNs to send and receive data over the public
network, but this makes them vulnerable to cyber-attacks.
The implementation of conventional cryptographic
algorithms is a very complex and computationally
intensive process. Employing these algorithms in WSNs
is a huge challenge since WSN nodes have limitations in
terms of computational power, storage memory,
communication bandwidth, and battery power/energy [4],
[8]–[13].
The biggest constraint in WSNs is energy. Work in [4]
suggests that energy consumption in WSN nodes can be
divided into three categories: (i) consumption by the
sensor transducer (to convert the physical quantities being
measured to electrical/electronic signals); (ii)
consumption by the communication module
(ZigBee/Lora/GSM etc.); and (iii) consumption during
microprocessor computation.
Authors of [14], [15] found that communication is more
energy consuming than cryptographic computation,
therefore message expansion as result of additional
information overheads introduced by cryptographic
algorithms come at a huge cost. Furthermore, the
implementation of enhanced security architectures leads
to more power consumption on the computation of
cryptographic functions. This implies that high level
security mechanisms introduce large communication and
computation overheads which then lead to high energy
consumption.
3. General Security Requirements for WSNs
The main goal behind implementing security in WSNs is
to protect the data that is being transmitted through the
network as well as the network resources against potential
attacks. In order to optimize conventional security
architectures for a given application, it is always essential
to be aware of the security requirements for that particular
application as it is the one that ultimately determines the
type of security architecture to be employed. The authors

of [4], [8] categorize common security requirements for
WSNs as described below:
Authentication: This is required to verify that the
communicating nodes are exactly who they claim to be
[4], [8]. It is very important for WSN nodes to have a
mechanism to confirm that the data they receive is indeed
from the actual trusted sender nodes. To encrypt data
without first being able to authenticate communicating
nodes is quite meaningless.
Confidentiality: This ensures that messages sent through
the network are unintelligible to all but the intended
recipient node [4], [16]. This maintains information
secrecy within the network.
Data Integrity: It ensures that the data received was not
altered or manipulated while in route from the source
node to the destination node [4].
Data Freshness: This ensures that the data received is
recent and not a replay of an old message [14].
Availability: This is meant to ensure that the services of a
WSN are always available and can be accessed even
during an attack [17].
Self-organization: It is essential for each node in a WSN
to be able to self-organize and self-heal. This poses a
challenge as it brings about the necessity for pre-key
distribution schemes to be employed [4].
Secure localization: This is required to securely get
accurate locations of sensor nodes in a WSN [4].
Time synchronization: Security mechanisms for WSNs
need to be time synchronized [4].
Various WSN applications normally focus on the
implementation of different security requirements
depending on the required security level, but the most
common are authentication, confidentiality, data integrity
and availability.
4. Security Attacks in WSNs
Security attacks affect a network’s capability and capacity
to perform its expected functions. It is imperative to
conduct a careful analysis of the various types of WSN
attacks in order to deduce possible countermeasures that
can prevent or minimize the associated effects. Authors of
[4] categorise WSN attacks into three groups: (i) attacks
on authentication and confidentiality; (ii) attacks on
service integrity; and (iii) attacks on network availability.
Table 1 is a summary of the most common WSN attacks
and their known countermeasures. Authors of [4], [8]
suggest that denial of service (DoS) attacks can be
analyzed effectively by classifying them according to the
layered network model. This approach identifies security
issues that each layer is susceptible to, and also allows
further analysis into attacks that can exploit the
interactions of the layers. Security attacks have a serious
impact on network performance and if left unattended
they may even render the network useless. It becomes less
of a challenge to propose effective security mechanisms
for WSN applications once security requirements and
associated security attacks have been identified and
thoroughly analyzed.
Table 1: Common WSN attacks and Associated Countermeasures as stated in [4], [8], [17], [26]
Attack Category
Types of Attacks
Possible Countermeasures
1. Attacks on authentication
and confidentiality
Eavesdropping, Traffic analysis,
Modification or spoofing of packets and
Packet replay attacks
Encryption and Authentication
2. Attacks on service
integrity
Compromised node used to feed the
network with false data values
Encryption, Hashing algorithms and
Authentication
3. Attacks on Network
Availability (DoS)
 Physical layer
Jamming
Spread spectrum and frequency hopping,
low duty cycles
Tampering
Tamper proof circuits and hardware
enclosures
 Data link layer
Collisions, unfair resource allocation
and resource exhaustion
Error correction coding, Time division
multiplexing, Rate limiting MAC
admission control
 Network layer
Spoofed routing information, selective
forwarding, Sinkhole, Sybil, Wormhole,
Hello Flood
Encryption, Authentication and Multipath
routing

5. Cryptographic Mechanisms for WSNs
Authors of [18] define cryptography as the science of
secret writing which is achieved through encryption.
Decryption is the process of data recovery in
cryptography.
A careful analysis and selection of the right cryptographic
mechanisms is fundamental to successful implementation
of optimized security architectures for WSN applications.
Most of the security services such as authentication,
confidentiality, integrity and non-repudiation are normally
ensured through the use of various forms of cryptography
and incorporating them into already existing but
simplified security protocols. The authors of [4], [8], [14],
[19] highlight the importance of evaluating cryptographic
algorithms with respect to storage size, operation speed,
data size and power consumption as a way of determining
their relative efficiencies. An algorithm’s efficiency can
further be evaluated by taking into account the security
requirements of the intended application and the
characteristic features (processing power, memory and
communication bandwidth) of a particular node under
consideration.
5.1 Evaluation of Symmetric Key Cryptography
Symmetric key cryptography uses the same key for both
encryption and decryption. Authors of [14] analyzed three
symmetric key algorithms; AES (Rijndael), RC5 and
RC6, and compared their energy consumption and
memory requirements on a Mica2 sensor mote. RC5
shows to be the most memory efficient, followed by RC6
and lastly AES. AES outperforms both RC5 and RC6
with regard to power consumption associated with the
computation of the cryptographic algorithms. Memory
efficiency has an impact on energy consumption because
energy is required to store data. However, computational
efficiency has far more significant energy cost
implications as compared to memory efficiency. Hence
the overall results show AES to be the most energy-
efficient symmetric cryptographic algorithm of the three.
Authors of [20] conducted similar work where they
compared and evaluated the energy consumption of three
symmetric key algorithms (RC4, RC5 and IDEA) and two
message digest/hash algorithms (SHA1 and MD5).
Experiments conducted were based on measuring
computational overheads of the respective algorithms on 6
different microcontroller platforms (Atmega 103, Atmega
128, SA-1110, UltraSparc2, M16C/10 and PXA250). The
experiments indicated mostly uniform computational
costs for the encryption algorithms, and it was also
observed that RC4 outperforms its successor algorithm,
RC5, in low end processors. Hashing algorithms (SHA1
and MD5) were observed to incur higher computational
overheads than cryptographic algorithms. Authors of [21]
evaluated the power consumption of encryption
algorithms (RC5, RC6, SkipJack, TEA and DES) on
Crossbow MICA2 sensor motes using TinySec. The
experiments took into account computational,
communication and memory implications on power
consumption. Their results showed that SkipJack and RC5
have better energy performance in WSNs, but SkipJack
however consumes more energy than RC5. Authors of
[22] compared the performance of AES and XXTEA to
the default TinySec algorithm, SkipJack, on MICA2
motes. Performance was evaluated based on CPU cycles,
throughput and power consumption, and experiments
showed XXTEA algorithm to be the most optimum for
WSNs. Authors of [23] studied and evaluated six block
ciphers (RC5, RC6, Rijndael, MISTY1, KASUMI and
Camellia) that according to literature are suitable
candidates for WSN applications. Experiments were
conducted on a 16-bit Texas Instruments microcontroller
MSP430F149, and the evaluation criterion took into
account security properties, memory and energy
efficiency of the selected algorithms. Results showed
Rijndael to be the best for high security and energy
efficiency and MISTY1 showed better performance in
storage and memory efficiency. Table 2 compares some
of the most common WSN symmetric ciphers by energy
efficiency on aspects of storage memory, processing
speed and communication.
Table 2: Ranking of symmetric ciphers by memory, processing and communication efficiency as evaluated from [21], [23]
Rank
Performance by
memory
efficiency (ROM)
Performance by
memory
efficiency (RAM)
Performance by
processing
efficiency
Performance by
message throughput
(with authentication
and encryption)
Performance by
communication latency
(with authentication and
encryption)
1
SkipJack, TEA
Rijndael
Rijndael, RC5
SkipJack, RC5
RC5
2
DES
RC5, RC6, TEA
SkipJack
TEA
SkipJack
3
RC5, RC6
SkipJack
TEA
TEA
4
Rijndael
RC6
5
DES
DES

5.1 Evaluation of Asymmetric Key Cryptography
Asymmetric cryptography is based on the use of two keys
that are mathematically related, one for encryption and the
other for decryption. The key pair is comprised of the
private and the public key. Each user has both keys, but
the private key remains a secret while the public key is
revealed to all other users. Asymmetric cryptography
incurs more computational overheads than symmetric
cryptography, but however simplifies the process of key
distribution and management as compared to symmetric
cryptography. This implies that asymmetric crypto-
systems are best suited for authentication and key
exchange services. According to [24], it costs 42mJ of
energy to encrypt a 1024-bit block on a MC68328
DragonBall processor using RSA, while the encryption
process of a 128-bit AES block is estimated to consume
much less at 0.104mJ.
In work [11], authors investigated and compared the
energy cost implications of authentication and key
exchange for two asymmetric algorithms RSA and ECC.
Experiments were performed on an 8-bit Atmel ATmega
128L low power microcontroller. Results show that ECC
has much better energy costs than RSA in both
authentication and key exchange processes. A similar
experiment was simulated in [25] to compare the energy
efficiency of ECC and RSA on MICA2DOT motes, and
the same results were obtained showing ECC as the better
choice of asymmetric cryptography for resource
constrained environments.
6. Conclusion
There exists no generic security solution for all WSNs.
Appropriate security architectures for WSNs greatly
depend on the security requirements of particular WSN
applications and hardware limitations of the type nodes
being employed.
Among the reviewed symmetric block ciphers, Rijndael
appears to be the most energy efficient and sufficiently
secure algorithm. Among the reviewed asymmetric block
ciphers, ECC has shown much better energy performance
while offering the same level of security as the most
commonly employed asymmetric algorithm RSA.
Selecting the right algorithm highly depends on
determining the most efficient, in terms of computation,
memory and energy, and sufficiently secure for a given
application.
This paper provides an overview of the energy
requirements of cryptographic algorithms when employed
in WSNs. This involved performing an evaluation of the
most commonly used symmetric and asymmetric
algorithms based on published literature, and developing
ranking model based on computational, memory and
bandwidth efficiency.
7. References
[1] D. Evans, “The Internet of Things - How the Next
Evolution of the Internet is Changing
Everything,” CISCO white paper, no. April. pp.
1–11, 2011.
[2] Lopez Research, “An Introduction to the Internet
of Things (IoT),” LOPEZ RESEARCH LLC, vol.
Part 1. of, no. November. pp. 1–6, 2013.
[3] L. Coetzee and J. Eksteen, “The Internet of
Things – Promise for the Future ? An
Introduction,” in IST-Africa Conference
Proceedings, 2011, pp. 1–9.
[4] J. Sen, “A Survey on Wireless Sensor Network
Security,” Int. J. Commun. Networks Inf. Secur.,
vol. 52, no. 2, p. 24, 2010.
[5] B. Manjuprasad and A. Dharani, “Simple Secure
Protocol for Wireless Sensor Networks,” in 2014
World Congress on Computing and
Communication Technologies, 2014, pp. 260–263.
[6] S. Maqbool and U. Sabeel, “Arising Issues In
Wireless Sensor Networks : Current Proposals
And Future Developments,” IOSR J. Comput.
Eng., vol. 8, no. 6, pp. 56–73, 2013.
[7] Libelium, “Libelium Website.” [Online].
Available: www.libelium.com.
[8] X. Chen, K. Makki, K. Yen, and N. Pissinou,
“Sensor network security: a survey,” IEEE
Commun. Surv. Tutorials, vol. 11, no. 2, pp. 52–
73, 2009.
[9] G. De Meulenaer, F. Gosset, F. X. Standaert, and
O. Pereira, “On the energy cost of communication
and cryptography in wireless sensor networks,” in
Proceedings - 4th IEEE International Conference
on Wireless and Mobile Computing, Networking
and Communication, WiMob 2008, 2008, pp.
580–585.
[10] W. Jiang, Z. Guo, Y. Ma, and N. Sang, “Research
on cryptographic algorithms for embedded real-
time systems: A perspective of measurement-
based analysis,” in IEEE 14th International
Conference on High Performance Computing and
Communications Research, 2012, pp. 1495–1501.
[11] A. S. Wander, N. Gura, H. Eberle, V. Gupta, and
S. C. Shantz, “Energy Analysis of Public-Key
Cryptography for Wireless Sensor Networks,” in
Third IEEE International Conference on
Pervasive Computing and Communications, 2005,
no. PerCom, pp. 324–328.

[12] K. Biswas, V. Muthukkumarasamy, E.
Sithirasenan, and K. Singh, “A Simple
Lightweight Encryption Scheme for Wireless
Sensor Networks,” in Distributed Computing and
Networking SE - 33, vol. 8314, Springer Berlin
Heidelberg, 2014, pp. 499–504.
[13] K. Biswas, “Lightweight Security Protocol for
Wireless Sensor Networks,” in A World of
Wireless, Mobile and Multimedia Networks
(WoWMoM), 2014 IEEE 15th International
Symposium on, 2014, pp. 1 – 2.
[14] S. Ben Othman, “Performance evaluation of
encryption algorithm for wireless sensor
networks,” in Information Technology and e-
Services (ICITeS), 2012 International Conference
on, 2012, pp. 1 – 8.
[15] G. Gaubatz, J. P. Kaps, E. Ozturk, and B. Sunar,
“State of the art in ultra-low power public key
cryptography for wireless sensor networks,” in
Pervasive Computing and Communications
Workshops, 2005. PerCom 2005 Workshops.
Third IEEE International Conference on, 2005,
pp. 146–150.
[16] H. Hayouni, M. Hamdi, and T.-H. Kim, “A
Survey on Encryption Schemes in Wireless
Sensor Networks,” in 2014 7th International
Conference on Advanced Software Engineering
and Its Applications, 2014, pp. 39–43.
[17] Y. Zhou, Y. Fang, and Y. Zhang, “Securing
wireless sensor networks: a survey,” IEEE
Commun. Surv. Tutorials, vol. 10, no. 3, pp. 6–28,
2008.
[18] C. Paar and J. Pelzl, Understanding
Cryptography. Springer Berlin Heidelberg, 2010.
[19] Y. Wang, G. Attebury, and B. Ramamurthy, “A
survey of security issues in wireless sensor
networks,” IEEE Commun. Surv. Tutorials, vol. 8,
no. 2, pp. 2–23, 2006.
[20] P. Ganesan, R. Venugopalan, P. Peddabachagari,
A. Dean, F. Mueller, and M. Sichitiu, “Analyzing
and modeling encryption overhead for sensor
network nodes,” in Proceedings of the 2nd ACM
international conference on Wireless sensor
networks and applications - WSNA ’03, 2003, p.
151.
[21] G. Guimaraes, E. Souto, D. Sadok, and J. Kelner,
“Evaluation of Security Mechanisms in Wireless
Sensor Networks,” in 2005 Systems
Communications (ICW’05), 2005, pp. 428–433.
[22] G. Jolly, M. C. Ku, P. Kokate, and M. Younis, “A
Low-Energy Key Management Protocol for
Wireless Sensor Networks,” in Eighth IEEE
International Symposium on Computers and
Communication (ISCC’03), 2003, pp. 1–6.
[23] Y. W. Law, J. Doumen, and P. Hartel, “Survey
and benchmark of block ciphers for wireless
sensor networks,” ACM Trans. Sens. Networks,
vol. 2, no. 1, pp. 65–93, 2006.
[24] D. Carman, P. Kruus, and B. Matt, “Constraints
and approaches for distributed sensor network
security (final),” DARPA Project report. NAI
Labs, pp. 1–139, 2000.
[25] F. Amin, a. H. Jahangir, and H. Rasifard,
“Analysis of public-key cryptography for wireless
sensor networks security,” Proc. World Acad. Sci.
Eng. Technol., vol. 43, no. July, pp. 530–535,
2008.
[26] M. M. Patel, “Security Attacks in Wireless Sensor
Networks : A Survey,” in International
Conference on Intelligent Systems and Signal
Processing (ISSP), 2013, pp. 329–333.