SECURITY TOPOLOGY IN WIRELESS SENSOR NETWORKS WITH

ROUTING OPTIMISATION

*M.Ismail and **M.Y.Sanavullah

*Research Scholar, Electronics and Communication, Vinayaga Mission University, Salem
**Professor and Dean, Dept. of EEE, K.S.R. Engg. College, Tiruchengode.
e-mail id: ravi_mls@yahoo.com

ABSTRACT
Multiple sensor nodes deployed in a common neighborhood to sense an event and subsequently transmit
sensed information to a remote processing unit or base station, has been the recent focus of research. Tiny sensor
nodes, which consist of sensing, data processing, and communicating components, leverage the idea of sensor
networks based on collaborative effort of a large number of nodes. These numerous sensors are used (similar to
different sensory organs in human beings) for delivering crucial information in real-time from environments and
processes, where data collection is impossible previously with wired sensors [1]. Typically, wireless sensor
networks are composed of low power sensor nodes and integrate general-purpose computing with heterogeneous
sensing and wireless communication. Their emergence has enabled observation of the physical world at an
unprecedented level of granularity. One of the most important components of a sensor node is the power unit and
may be supported in most applications by a power scavenging unit such as solar cells. Hence, there is a major
limitation in a wireless sensor networks, such as, the sensor nodes must consume extremely low power. Also,
wireless networks are subject to various kinds of attacks and wireless communication links can be eavesdropped on
without noticeable effort and communication protocols on all layers are vulnerable to specific attacks. In contrast to
wire-line networks, known attacks like masquerading, man-in-the-middle, and replaying of messages can easily be
carried out. Hence, a fundamental issue in the design of wireless sensor networks is the reliability i.e. how long can
the wireless sensor networks survive and how well are the wireless sensor networks recovery after the malicious
attacks. In this context, in this thesis, the power, mobility, and task management planes that can monitor the power,
movement, and task distribution among the sensor nodes are proposed. These planes help the sensor nodes
coordinate the sensing task and also lower the overall power consumption. In addition, a secure topology discovery
algorithm is proposed and its performance is studied for different types of node distributions. The proposed work is
the development of architecture for secure communication in mobile wireless networks. The approach divides the
network into clusters and implements a decentralized certification authority. Decentralization is achieved using
threshold cryptography and a network secret that is distributed over a number of nodes. While this basic idea has
been proposed earlier partially, its application on a clustered network is a novelty.

1. INTRODUCTION
A sensor node is made up of four basic components
as shown in figure 1: a sensing unit, a processing
unit, a transceiver unit and a power unit [5]. They
may also have application dependent additional
components such as a location finding system, a
power generator and a mobilizer. One of the most
important components of a sensor node is the power
unit and may be supported in some cases even by a
power scavenging unit such as solar cells. These
nodes must
(i) consume extremely low power
(ii) operate in high volumetric densities [12]
(iii) have low production cost and be dispensable and
(iv) be adaptive to the environment.
The primary objective of this paper is (i) to propose a
secure topology discovery algorithm for a WSN (ii)
To optimize the routing table and effectively route
the data in the network layer and to design a power
aware Medium Access Control (MAC) protocol
(where the environment is noisy and sensor nodes
can be mobile [7]), (iii) to minimize collision with
neighbors broadcast and to effectively detect the
aberrant nodes and eliminate them (iv)To devise
simple but robust modulation, transmission and
receiving techniques for the physical layer and
provide WSNs the flexibility to balance detection
accuracy, sensor density and energy consumption.
Such energy efficient WSNs will extend the system’s
life time as sensor nodes usually rely on small and
non-renewable batteries [9].

978-1-4244-3328-5/08/$25.00 ©2008 IEEE 7

Authorized licensed use limited to: Korea Advanced Institute of Science and Technology. Downloaded on August 26, 2009 at 02:08 from IEEE Xplore. Restrictions apply.

Figure 1 Components of a sensor node.
2. SECURE TOPOLOGY DISCOVERY
ALGORITHM
The secure topology algorithm proposed in
this work, belongs to the class of prevention type
security. The malicious nodes are prevented from
eaves dropping while discovering the routes. This is
implemented by providing a secret key to all sensor
nodes and all the route information messages are
encrypted. Perimeter security is the application
chosen to illustrate the security protocol. The
following assumptions are made in this work: (i) The
base station is computationally robust, and possess
the requisite processor speed, memory and power to
support the cryptographic and routing requirements
of the sensor network. (ii)The base station is part of a
trusted computing environment. (iii) The
communication paradigm is either base station to
sensor or vice-versa (iv)The radio range of a sensor is
15 meters and sensing range is 1m. (v) The protocol
provides for a multi-hop scenario where the range of
a base station is extended employing nodes that are
adjacent to the base station to serve as intermediaries
for non-adjacent nodes.
2.1 SINGLE COLLECTION AND
AUTHENTICATION POINT (BASE STATION)
MODEL
Consider the family of sensor routing
protocols where each sensor communicates either
directly or indirectly with a base station. In turn the
base station correlates and aggregates information
from each sensor. Accordingly, the base station will
need to verify the authenticity of the sensor, the
integrity of the communication and ascertain that it is
not a replay of an earlier communication. Recall the
assumption that the base station is computationally
robust and secure. In this security protocol each
sensor j shares a unique 64 bit Key Keyj with the base
station. The protocol provides for a multi-hop
scenario where the range of a base station is extended
Authorized licensed use limited to: Korea Advanced Institute of Science and Technology. Downloaded on August 26, 2009 at 02:08 from IEEE Xplore. Restrictions apply.
employing nodes that are adjacent to the base station
to serve as intermediaries for non-adjacent nodes.
Figure 2 depicts an example of such a sensor network
topology.
Figure 2 Example Network Topology
The format of all communication (sensor
nodes and the base station) consists of a preamble,
header and payload. The preamble is empty if the
communication originates from the base station and
is directed to a sensor; otherwise it contains the
address of the sending node. The header contains the
recipient’s address, nonce and a command and is
encrypted under key Kj, which is shared between the
base station and node j. The payload contains data
exchanged between the node and the base station. As
will be explained, the payload is encrypted under the
shared key of the destination node, which may be
different from the key used to encrypt the header.
This difference comes into play when the
communication needs to be relayed by an
intermediate node. Figure 3 depicts the
communication format. The size of the packet is
assumed to be 40 bytes.
Figure 3 Message Format
2.2 DISCOVERY ALGORITHM
The base station is deployed with the unique
ID and symmetric encryption key of each node in the
micro sensor network. Similarly, each node is
deployed with the unique key that it shares with the
base station and its clock is synchronized with the
base station’s clock.
2.2.1 Adjacent node Discovery
A node is called an adjacent node if it is
within the broadcast range of the base station. To
 discover a node as adjacent node or not, the the base station’s address to the preamble
following steps are performed: and the Ψ to the payload (Fig. 6 Step 3).
1. The base station sends a HELLO message 4. In turn, the adjacent node receives the
to each node. transmission, decrypts the header and upon
2. If the node replies with a HELLO-REPLY, seeing the RELAY command, adds the
then the node is adjacent to the base station and the preamble to the payload and transmits it to
base station adds that node to its route table. This is the base station (Fig. 6 Step 4).
illustrated in fig. 4. 5. The base station after receiving the HELLO-
REPLY adds the adjacent node as one of the
route to reach the non-adjacent node.

Figure 4 Adjacent node discovery

2.2.2 Non-adjacent node discovery

A non-adjacent node is one which is not

reachable directly by the base station (Figure 5). To
discover the non-adjacent node, the base station uses
the adjacent nodes. The base station tries all the
adjacent nodes to reach the non-adjacent node. The
adjacent nodes which are used to reach the non-
adjacent node are noted as the route to reach the non-
adjacent node.
Figure 6 Non-adjacent node discovery
Algorithm

After performing the secure topology

discovery, the base station contains a route table that
Figure 5 Network model represents a constituent of nodes and their route to
reach these nodes. In this table the route to adjacent
The detail implementation procedure is node is mentioned as direct. This route table is called
given below. as raw route table since it has many redundant routes
1. The base station sends a message containing to reach a non-adjacent node. The purpose of route
the RELAY command and a payload, to be table optimization is to assign impartial load to all
forwarded to the non-adjacent node, to each adjacent nodes [6].
of the adjacent nodes (Fig. 6 Step 1).
2. The relaying (adjacent) node adds the 3. ROUTE TABLE OPTIMIZATION
original header to the payload (which ALGORITHM
contains Ψ) and transmits the new message
to the non-adjacent node which contains the For a non-adjacent node, the route table
HELLO command (Fig. 6 Step 2). optimization algorithm involves calculation of load
3. To respond to the HELLO message, the non- for each adjacent node and the node with minimum
adjacent node constructs a HELLO-REPLY load is assigned as the route to the non-adjacent node.
message encrypting it under the key it shares The algorithm involves the following steps:
with the base station and places it in the For a non-adjacent node
payload. The message is transmitted adding 1. Calculate the load for each adjacent node.

Authorized licensed use limited to: Korea Advanced Institute of Science and Technology. Downloaded on August 26, 2009 at 02:08 from IEEE Xplore. Restrictions apply.
 2. Chose minimum load adjacent node as the
route.
3. Steps 1 and 2 are repeated till all the non-
adjacent nodes are assigned routes.
This is given in appendix-1 of this paper.
3.1 Load calculation
Load is calculated based on the number of
past assigned nodes (confirmed) and the number of
future possible assignment of nodes. Since the past
assigned nodes are known, they are given higher
weightage than the future possible assignments. For
an adjacent node the load calculation is done using
the formula
Load = mP + nQ
Where ‘P’ is the number of past assigned non-
adjacent nodes to the adjacent node and ‘Q’ is the
number of future possible assignment of non-adjacent
nodes to the adjacent node, ‘m’ and ‘n’ are the
weights for P and Q respectively.
3.2 Optimal weight
In this algorithm the optimal weight is
chosen by trial and error method. The optimized
route table is compared for different weights and the
set of weights which results in best route table are
chosen as the optimal weights. In this work, by
simulation the optimal weights are found to be m = 1
and n = 0.5. The simulation results are presented in
the last section.
taken for simulation etc. The network animator was
used to view the simulation in its animated form. To
check the validity of the algorithms proposed in this
work, different test inputs were given and the results
were analyzed. The secure topology discovery
algorithm was validated with random distributions
like Poisson, Rayleigh and Exponential. The various
results obtained are explained in the following
section.
5.1 DISCUSSION
It was assumed that 12mA of current was
drawn to transmit a message and 1.8mA to receive a
message. The base station was located at the centre
and 24 sensor nodes are randomly distributed around
the base station. The secure topology discovery
algorithm is validated by giving various random
distributions as the input. The distribution of sensor
nodes is made with reference to the base station
which is located at the centre. The distributions
considered for study in this work are Poisson,
Rayleigh and exponential probability distribution
function. A known placement of nodes (uniform
distribution of nodes) is also considered for study.
These are all shown in figures 7 to 10.

4. CIPHER UPDATING

Cipher updating is the process by which the

non-adjacent nodes are given a payload to enable
communication with the base station. The primary
route is the cipher given to the non-adjacent nodes.
The Cipher contains the address of the primary route
node, a nonce and COMMAND encrypted by the key
of the primary route node. An example cipher can be
given by considering the Table 4. Cipher for the node
E will be like Ψ = KB {Address of B, DTG, RELAY}

5. RESULTS AND DISCUSSION

The secure topology discovery algorithm of Figure 7 Known Placement of nodes

section 2 was simulated in ns2 (network simulator 2) .
and the analysis graphs are plotted using MATLAB.
The route table optimization algorithm of section 3
was implemented in MATLAB. The neglect type
DoS threat was simulated in ns2. Ns2 used 914MHz
license free ISM (Industrial, Scientific and Medical)
band as the frequency of operation. The output of ns2
is used to compute the energy consumption, time

10

Authorized licensed use limited to: Korea Advanced Institute of Science and Technology. Downloaded on August 26, 2009 at 02:08 from IEEE Xplore. Restrictions apply.
 Figure 10 Poisson distribution of sensor nodes
Figure 8 Exponential distribution of sensor nodes
In this work the random node distribution
that gives the best performance for the proposed
algorithm is evaluated. It is established that Poisson
distribution of nodes gives best performance in terms
of time taken and energy required to discover the
topology. The route table optimization algorithm is
implemented for a sample node distribution and the
results are presented. Fig.11 to 13 shows the load on
nodes. The comparison of these plots proves that the
value of weight n = 0.5 results in equal share of load
to the nodes.

Figure 9 Rayleigh distribution of sensor nodes

The plots depict the placement of nodes in

the actual field in terms of the X and Y coordinates of
the nodes. Figure 11 Load on nodes for n = 0.3

11

Authorized licensed use limited to: Korea Advanced Institute of Science and Technology. Downloaded on August 26, 2009 at 02:08 from IEEE Xplore. Restrictions apply.
 Figure 15 shows the energy spent in transmission and
reception to discover the given topology using the
secure topology discovery algorithm. Figure 16
shows the energy spent in transmission comparing
the various distributions. The throughput fluctuation
plot obtained is given in figure 17.

Figure 12 Load on nodes for n = 0.5

Figure 15 Energy consumption for topology

discovery

Figure 13 Load on nodes for n = 0.7

Figure 14 gives the screen shot of simulation

results in ns2 for neglect type DoS threat. In figure,
the blue colored node is the base station. The yellow
one is the malicious node and the black ones are the
sensor nodes that are not triggered. The green colored
nodes are the nodes triggered by the phenomenon
depicted by red color.

Figure 16 Energy spent in transmission for

different random node distributions
Figure 14 ns2 screen shot for neglect type
DoS threat

12

Authorized licensed use limited to: Korea Advanced Institute of Science and Technology. Downloaded on August 26, 2009 at 02:08 from IEEE Xplore. Restrictions apply.
 (5)D. Culler, D. Estrin and M. Srivastave, “Overview of sensor
networks”, IEEE Computer, 37: 4149, August 2004.
(6)J.Elson and D.Estrin,“ Wireless sensor networks”,Kluwer
Academic Publishers, 2004, Ch. Wireless sensor networks: A
bridge to the physical world.
(7)Standard for part 15.4: Wireless Medium Access Control
(MAC) and Physical layer (PHY) specifications for low rate
wireless personal area networks (WPAN), IEEE std 802.15.4,
IEEE, Newyork 2003.
(8)J.Polastre, J.Hill, D. Culler, “Versatile low power media access
for wireless sensor networks”, Proceedings of the SenSys04,35,
November, 2004, Baltimore, Maryland.
(9)Lige Yu, Lin Yuan, Gang Qu, Anthony Ephremides, “ Energy-
driven detection scheme with guaranteed accuracy”, IPSN 06,
April 19-21, 2006, Tennessee, USA.
(10)Wenyu CAI, Xinyu Jin, Yu Zhang, Kangsheng Chen, Jun
Tang, “Research on reliability model of large scale wireless sensor
networks”,1-4244-0517-3/06, 2006 IEEE.
(11)M. Cardei and D.Z. Du, “Improving wireless sensor network
lifetime through power aware organization”,ACM wireless
networks, Vol. 11, No.3, 2005.
(12)F. Akyilidz, W.Su, Y. Sankarasubramaniam and E.Cayirici, “
A survey on sensor networks”, IEEE communication magazine,
Aug 2002: 102-114.

8. BIOGRAPHIES

(1)Mr. Mohammed Ismail is presently a research scholar in

Figure 17 Comparison plot of throughput Vinayaka Mission University, Salem, and working as
fluctuation for Rayleigh, Exponential and known Associate Professor in department of ECE, SR Engineering
distribution College, Warangal. His areas of interest include wireless
sensor networks, fault tolerant networks, etc.

6. CONCLUSION

The present work is ideally suited to most

of the applications that share similar features such as
difficult to access because of geographical location
where the network has been deployed, the large scale
of deployment, high mobility and prone to failure. It
also ensures that the WSN be autonomous and
operate unattended, be adaptive to the environment
and choose an optimal number of communicating (2) Dr. M.Y.Sanavullah is presently Dean, Faculty of
sensing nodes since too many sensors can generate Electrical Engg.. His areas of interest include fault tolerant
bottlenecks in the communication infrastructure when system design, sensor signal processing and evolvable
they all compete for bandwidth. hardware.

7. REFERENCES

(1) Ruizhong Lin, Zhi Wang and Youxian Sun, “ Wireless sensor
networks solutions for real time monitoring of nuclear power
plant” , In Proceedings of 5th world congress on Intelligent Control
and Automation, June 15-19, 2004, China.
(2)Hairong Qi, Phani Teja Kuruganti and Yingyue Xu, “The
development of localized algorithms in wireless sensor
networks”,Computer Networks, 2002, 38, 286-292.
(3)Yue-Shan Chang, Chih-Jen Lo, Ming-Tsung Hsu, Jiun-Hua
Huang and Tong-Ying Juang, “ Fault estimation and fault map
construction on cluster-based wireless sensor network”,
Proceedings of the IEEE international conference on Sensor
networks, Ubiquitous and Trustworthy computing, 2006.
(4)Bonnie Zhu, Bruno Sinopoli, Kameshwar Poolla and Shankar
Sastry, “Estimation over wireless sensor networks”, Proceedings of
the 2007 American Control conference, July 11-13, 2007.

13

Authorized licensed use limited to: Korea Advanced Institute of Science and Technology. Downloaded on August 26, 2009 at 02:08 from IEEE Xplore. Restrictions apply.
 APPENDIX-I
Table 1 Connection table

1 2 5 7 10 11 13 14 15 16 17 19 20 21 24
3 0 0 0 0 0 0 1 0 0 1 1 0 1 0 1
4 1 0 0 0 0 0 1 0 1 0 0 0 0 1 0
6 1 0 0 1 0 0 1 0 1 0 0 1 0 1 0
8 0 0 0 1 0 0 0 0 0 0 0 1 0 1 0
9 0 0 0 0 0 0 0 0 0 1 1 0 0 0 1
12 1 0 0 0 0 0 1 0 1 0 1 0 1 0 0
18 1 1 0 1 0 0 1 0 1 0 0 1 0 1 0
22 0 0 1 0 0 1 0 0 0 1 0 0 0 0 1
23 1 1 0 1 1 0 0 0 0 0 0 1 0 1 0

Table 2 Primary route Table

1 2 5 7 10 11 13 14 15 16 17 19 20 21 24
3 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0
4 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0
6 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0
8 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0
9 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0
12 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0
18 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0
22 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0
23 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0

Table 3 Secondary route Table

1 2 5 7 10 11 13 14 15 16 17 19 20 21 24
3 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0
4 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0
6 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0
8 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0
9 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0
12 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0
18 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0
22 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0
23 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0

Note: Table 1 shows the connection table. Table 2 and 3 are obtained after applying the optimization algorithm. The
optimized route table is given in Table 4.

14

Authorized licensed use limited to: Korea Advanced Institute of Science and Technology. Downloaded on August 26, 2009 at 02:08 from IEEE Xplore. Restrictions apply.
 Table 4 Optimized route Table

Node Primary Route Secondary

Route
1 D

2 D

3 20 16

4 13 15

5 D

6 15 1

7 D

8 7 19

9 16 17

10 D

11 D

12 17 20

13 D

14 D

15 D

16 D

17 D

18 1 2

19 D

20 D

21 D

22 5 11

23 2 7

24 D

15

Authorized licensed use limited to: Korea Advanced Institute of Science and Technology. Downloaded on August 26, 2009 at 02:08 from IEEE Xplore. Restrictions apply.