A new secure localization approach of wireless sensor nodes in the presence of misbehaving anchor nodes

International Journal of Computer Science, Engineering and Information Technology (IJCSEIT), Vol. 4, No.3, June 2014
DOI : 10.5121/ijcseit.2014.4301 1
A NEW SECURE LOCALIZATION APPROACH OF
WIRELESS SENSOR NODES INTHE PRESENCE OF
MISBEHAVING ANCHOR NODES
ZohrehSobhanifar1
and AbolfazlToroghiHaghighat2
1,2
Department of Electrical, Computer and Biomedical Engineering, Qazvin branch,
Islamic Azad University, Qazvin, Iran
ABSTRACT
This paper proposes a new algorithm to find and isolate the nodes which lies about their position in a
Wireless Sensor Network (WSN). Also the proposed method enables the sensor nodes to find their location
in presence of liar nodes. In the proposed method, a given number of neighbors for all sensors is
determined wherethe number of liars is below a predefined threshold value. The proposed method is
evaluated in finding the liars and also the correct location of each node. The minimum error rate on the
determination of liars and the location of sensors proves the ability of the algorithm for localization of
sensors in the WSNs.
KEYWORDS
Wireless Sensor Network (WSN),Secure Localization,Wireless Security, Network Security
1. INTRODUCTION
Wireless Sensor Networks (WSNs) provides many shared activities among all users where the
activities are considered as a small part of pervasive technology. They are used in many
interesting applications. Realization of these applications requires wireless ad hoc networking
techniques [1].The applications of WSNs in both military and civilian cases reveal the necessity
of wireless ad hoc networking such as target tracking, environmental monitoring,and trafﬁc
regulation, to name just a few [2].
WSNs are build up by deploying multiple microtransceivers called sensor nodes that allow users
to gather and transmit data from regions which might be inaccessible or hostile to human beings.
The data transmission is performed independently by each node in a wireless medium. All sensors
in WSNs use multi-hop communications where they collaborate with each other to route collected
data to the target [3]. In WSNs, routing algorithms is based on different measures such as network
structure or the protocol operations [1]. In the routing protocol, the sensor node positions are
utilized to route the packet data in the network. In the geographical area, the information of all
sensors must be collected in order to locate their positions.
The location of sensors must be secured to provide the integrity of the WSN and its associated
services. At the first step, the necessary parameters for each node are determined to find all paths
in order to lead their data towards the targets [4].The information of node’s positions is utilized to
analyse the data collected by the sensors, i.e. the origin of collected data should be known for the

2
user before using it. Finally, when the end user needs to collect information from some nodes, it
should be send the position the query nodes. The localization process is therefore crucial [5].Note
that the false information on the geography of the phenomenonprovided by attacker is received by
sensors. To secure the localization of WSN, most of the work is based on the use of trust models
when the trusted nodescalled anchors provide information to localization processes (e.g.
localization process using several GPS-based anchors [3].
A new class of anchor nodes is verifiers which review the information provided by the anchors.
This model is often too expensive and not always realistic. Firstly, to ensure the coverage of
network, the distribution of anchors and veriﬁers should be determined a priori. The
representation of these special nodesis likely to be inferior because their cost is higher than the
cost of regular nodes [6]. On the other hand, the cryptographic approach is necessary to apply
trust on WSNs. It is important that the localization process should have a minimum effect on the
energy consumption.Finally, to deploy the integrity of sensors, the trusted nodes should be
monitored and so too much trust may decrease the independently of a WSN [3].
In this paper, the goal is the providing the information bout the location of regular nodes in the
presence of neighboring sensors which may lie about their location.In this paper, it is considered
that the liar sensors known their locations,but that, for any reason, inform false position to their
neighbors. Their reasons may be malicious(i.e., to mislead regular sensors into wrong locations)or
inadvertent (i.e., due to the presence of obstacles that prevent them from providing correct
locations). The proposed method, it is considered that the position of regular nodes is determined
in the situation that the number of liar nodes in the neighborhood of each regular node is below a
predefined threshold value. The proposed method enables regular nodes to find and isolate the liar
nodes.
2. LOCALIZATION IN THE PRESENCE OF LIARS
Assume that the location of the specific node A should be determined. The point with location
(a ,a )is the intersection of three circles which arecentred atB ‫و‬ B ‫و‬B and with radii
d(A, B ), d(A, B )and d(A, B ), respectively. Note that the locations of B ‫و‬ B ‫و‬B are
B (b ,b )‫و‬B (b ,b ) and B (b ,b ), respectively [3-4,7].
Figure 1.Localization of sensor A using three nodes B1, B2, and B3 that are located in its distance one
neighborhood (Captured from [3]).
The function d is calculated using following equations [3]:

3
( − ) − − = ( , ) (1)
(b − a ) − b − a = d(A, B ) (2)
( − ) − − = ( , ) (3)
Consider now that the sensor A may receive signals from the nodes which announce incorrect
locations to A. Assume that (A)shows the set of neighborhoodsensors which is located at the
distance one hop away from A, and lrefers to the number of liar nodes(l≤ (A)) (see Figure2).
Let us that the liar is parameterised with S. In this paper, a method is described to findthe
correctposition of regular nodes in the presence of liar nodes. Also, the proposed algorithm is able
to exclude the incorrect and at isolating the set of liars. Note that A knows the upper bound l of
sensor nodes lying in the geographical area. The proposed method is analyzedfor a given number
of liars.
In Section 3, the proposed algorithm uses the pairs of signals received from two nearest neighbor
and also a majority rule. We assume that , ∈ (A) is the pair of points in the neighbor of
Awhich computes the pairs of points ( , ),( ′ , ′ ) as the intersection of two circlescentered
at , with radii ( , ) and ( , ), respectively. Note that the majority rule is utilized to
correct location of A.We show that if the numberof liars among the nodes in (A) is higher than
one, A isable to determine its proper location if >
√
.
3. SECURE LOCALIZATION USING TWO NEIGHBOR SIGNALS
Let us sensor A uses the signals of two neighbors and so the correct location is one the two points
of intersection of the two circles centered at these two neighbours [11]. In this method, for every
pair of neighbors , ∈ (A), sensor Acomputes a pair of locations {x,x'} which is the
intersection of two circles with centers , .
As the shown inFigure 2, the proper location of A is either x or x'. After that, A may use the
majority rule to find the best position for A and to report liar nodes. It is important that at least
one of the nodes in all possible pairs ofthem is liar. In this regard, this paper can have either [3,8]:
1- both sensors are liars, for a total of
2
pairs,or
2- exactly one sensor is liars for a total of
−
1
.
1
pairs.
If a pair of locations is determined by two truth tellers, this pair is correct, otherwise it is
incorrect. The majority rule is utilized to accept the best position [9]. The rule is accepted if the
number of incorrect location is lower than the number of correctpairs of locations. This amounts
to having the inequality [10,12]:
2
−
2
−
−
1
.
1
>
2
+
−
1
.
1
(4)
from which we derive the inequality:

4
2
> 2
2
+
−
1
.
1
(5)
Table 1.Relation between the minimum number of liars and the minimum number of neighborsto determine
a correct pair of locations (using Algorithm 2) for a specific node.
As the previous mentioned, the proposed method works for a given number of liars. In this
regard, the minimum number of neighbor for a given number of l liars is presented in Table2. To
prove the Table 2, reader can refer to [3-4].
Let us that the number of neighbors andliars characterized with n and l, respectively.Assume that
= 5and = 1, the proposed method can be used to find a correct pair of locations, {x,x'}. Since
there is 5 neighborsfor A, from which one sensor is liar, the remaining sensors are truth tellers. To
find the liar node, following approach is proposed: at first, all intersection nodes{x,x'} of all
possible pairs of 5 nodes are produced. After that the pair of sensor which produces two
intersection nodes with maximum distance from the other intersection nodes, are selected as the
possibly liar nodes (suspicious pair).The next step is analyzing of each of sensor of the suspicious
pair in the other pairs of these sensors. Finally, the sensor that in all its pairs, most of the
produced intersection nodes have the maximum distance to the other intersection nodes is chosen
as the final liar node. The liar node is removed from 5 neighbor are truth tellers. After removing
of the liar node, the task is determination of actual location of A. For this purpose, the sensor
where the average of its intersection node’s locations (local average) has the nearest distance to
the average of all intersection node’s locations (global average), is chosen as the actual location
of A. It means that the selected node which produces the best intersection nodes is considered as
the best choice. At the final stage, the majority rule is employed to select one of the intersection
points of the best sensor, {x,x’}, as the actual location of A.
It is clear that the distribution of intersection nodes is the criterion of secure localization which
consists of finding the correct location of a specific sensor and reporting the liars. It is reasonable
that the node with maximum distance to other nodes is considered as the liar node and the node
with the minimum distance to the mean of all intersection nodes is considered as the correct
location a specific node.
4.SIMULATION
In this section, the proposed method is evaluated to prove the ability of the method to allow their
sensors to find their own location in an arbitrary WSN under the presence of liars. For this
purpose, a uniform distribution of n sensors located randomly and independentlyin the interior of
a unit square is considered. Note that in this simulation, n is set from twenty to 100-sensors and
the number of liars is the percentage of all sensors. The evaluation step is divided into two parts,
Number of liars Minimum number of neighbors
1 5
2 7
3 9
4 11
5 13
10 23
15 33
20 43

5
one part is about the ability of the proposed method in liar detection and the second one is about
the evaluation of the proposed method in finding the correct location of each node. Results of the
proposed method is compared the basis algorithm [3]. The difference between the basis algorithm
[3] and the proposed method is that in the basis algorithm [3], distribution of the intersection
nodes is not considered which is effective in the localization process.
In Figure2, the error rate on the detection of liar nodes is demonstrated in different number of
nodes, [20,100], using the following equation
Error Rate =
.
.
(6)
In the above equation, the error rate is computed on the detected liars which are incorrect. This
simulation tests the ability of the proposed method in finding actual liars. According to the figure,
maximum value is obtained in the size if twenty nodes (minimum size). Note that in each number
of nodes, 30% of all sensor are liars. This figure proves that the number of liar nodes is increased
with increasing number of nodes while the error rate on the liar node detection is not increased
and so, the proposed algorithm is robust to increasing number of liars.
Figure 2.Error rate on the detection of liar nodes in different number of nodes.
In Table 2, the results of error rate on liar detection are reported in all number of liars. In this
table, the minimum number of neighbors for each number of liars is determined. In each case, the
average value is the mean of error rate during 10 iterations. In this simulation, the average error
rate is increased along with increasing in the both of the number of liars and the minimum
number of neighbors. In comparison to Algorithm 2 [3], the best results is achieved using the
proposed method because of the idea of distribution of sensors. In the other words, the pair of
intersection nodes with maximum distance to the other intersection nodes is reasonable to
consider as the liars.
In Figure 3, the percentage of correct localization of all nodes is shown. In this regard, each node
unaware of its location requests the location of its neighbors. The simulation was run for 100
times for each network size. Results of the simulation for the networks with size of twenty to 100-
sensors aredemonstratedinFig. 3 which provesthe ability of the method to increase the number of
sensors that aware of their position when the average of liars is low.

6
Figure 3.Percentage of correct localization of all nodes in different number of nodes.
Note that in this simulation, the proposed method is evaluated in the case which the number of
liars is decreased with increasing of network size. It is important that the minimum number of
liars is 50% of total network size. It is important that the percentage of position aware nodes is
improved in the proposed method. The reason of superiority of the proposed method to the other
algorithm refers to the idea of distance between the intersection nodes and mean of all
intersection node locations. According to this idea, the intersection nodes which are near to the
mean of all intersection nodes are reasonable to be the correct location of a specific node.
5. CONCLUSION
This paper presents a new methodology to deal the localization process of WSN nodes in the
presence of liars. The proposed method has two properties: providing of correct position of each
sensor, and finding and isolation of the liar nodes.The proposed method utilizes the distribution of
intersection nodes where the intersection nodes with maximum distance to the other nodes can be
considered as the liars and the intersection node with minimum distance to mean of all
Table 2. Simulation results in different number of liars obtained using the Alg. 2 [3] and the proposed
method
Case
Number of
liars
Minimum number of
neighbors
Average of error rate
Alg. 2 [3] Proposed method
1 1 5 18.17% 15.87%
2 2 7 20.21% 16.91%
3 3 9 21.98% 17.99%
4 4 11 23.50% 19.10%
5 5 13 24.65% 20.36%
6 10 23 25.13% 21.85%
7 15 33 27.18% 23.24%
8 20 43 28.91% 24.79%

7
intersection nodes is considered as the correct location of a specific sensor. The simulation results
show the best results in finding the liars and the correct location of each node.
REFERENCES
[1] Al-Karaki, J. N. &Kamal, A. E. (2004)“Routing techniques in wireless sensor networks: A
survey”,IEEE Wireless Communications, Vol. 11, No. 6, pp. 6–28.
[2] Barbeau, M., Kranakis, E., Krizanc, D. & Morin, P. (2004) “Improving Distance Based Geographic
Location Techniques in Sensor Networks”, 3rd International Conference on AD-HOC Networks &
Wireless (ADHOC-NOW’04), pp. 197–210, LNCS 3158, Springer.
[3] Alfaro J.G., Barbeau M. &Kranakis E. (2009) “Secure localization of nodes in wireless sensor
networks with limited number of truth tellers”. In: 7th annual communication networks and services
research (CNSR) conference. IEEE, Piscataway, NJ, pp. 86–93.
[4]Alfaro, J.G., Barbeau, M. &Kranakis E. (2011) “Secure Geolocalization of Wireless Sensor Nodes
In The Presence of Misbehaving Anchor Nodes”, Annals of Telecommunications-Annales des
telecommunications, Vol. 66, No. 9-10, pp. 535-552.
[5] Bahl P., Padmanabhan V.N. &Balachandran A. (2000) “Enhancements to the RADAR user location
and tracking system”, Microsoft Research, Technical Report MSR-TR-2000-12, Microsoft, p 13.
[6] Capkun, S. &Hubaux JP (2005) “Secure positioning of wireless devices with application to sensor
networks. In: 24th annual conference of the IEEE computer and communications societies, Vol. 3.
IEEE, Piscataway, NJ, pp. 1917–1928.
[7] Capkun, S, Cagalj, M, Srivastava, M. (2006) “Secure localization with hidden and mobile base
stations”. In: 25th annual conference of the IEEE computer and communications societies. IEEE,
Piscataway, NJ, pp. 1–10.
[8] Capkun, S. &Hubaux J.P. (2006) “Secure positioning in wireless networks”,IEEE J Sel Areas
Commun,Vol. 24, No. 2, pp.221–232.
[9] Delaet S., Mandal P., Rokicki M. &Tixeuil S. (2008) “Deterministic secure positioning in wireless
sensor networks”, In: IEEE international conference on distributed computing in sensor networks
(DCOSS). Springer, Berlin, pp 469–477
[10] Doherty, L. &Ghaoui L.E. (2002) “Convex position estimation in wireless sensor networks”, In: 20th
annual joint conference of the IEEE computer and communications societies. IEEE, Piscataway, NJ,
pp. 1655–1663.
[11] He T., Huang C., Blum B.M., Stankovic, J.A., Abdelzaher, T. (2003) “Range-free localization
schemes for large scale sensor networks”, In: 9th annual international conference on mobile
computing and networking. ACM, New York, NY, pp. 81–95.
[12] Hwang J., He T. & Kim Y. (2008) “Secure localization with phantom node detection”, Ad Hoc
Networks Vol. 6, No. 7, pp.1031–1050.
[13] Ji, X. &Zha, H. (2004) “Sensor positioning in wireless ad-hoc sensor networks using
multidimensional scaling”, In: 23rd annual joint conference of the IEEE computer and
communications societies. IEEE, Piscataway, NJ, pp. 2652–2661.

A new secure localization approach of wireless sensor nodes in the presence of misbehaving anchor nodes

Recommended

More Related Content

What's hot (15)

Similar to A new secure localization approach of wireless sensor nodes in the presence of misbehaving anchor nodes (20)

Recently uploaded (20)

A new secure localization approach of wireless sensor nodes in the presence of misbehaving anchor nodes