International conference on Communication and Signal Processing, April 3-5, 2013, India
A Cluster Based Intrusion Detection and
Prevention Technique for Misdirection Attack
inside WSN
Roshan Singh Sachan, Student Member, IEEE, Mohammad Wazid, Student Member, IEEE, Avita
Katal, Student Member, IEEE, D P Singh, R H Goudar
 sensitive data is communicated to the destination node an
Abstract— Constrained resources, the ad-hoc nature of
deployment and the vulnerability of wireless media are some of
the most challenging characteristics of Wireless Sensor Network
(WSN) which pose a need for unique security solutions. WSNs are
susceptible to various attacks, in which Misdirection a kind of
Denial of Service (DoS) attack is very difficult to detect and
defend. In Misdirection Attack, the intruder misdirects the
incoming packets to a node other than the intended node which
introduces high end- to- end delay (sometimes infinite) in the
network. The performance of the network (i.e. throughput) is also
degraded. Thus the detection and prevention of this attack
becomes very crucial. In this paper we have proposed a novel
Cluster Based Intrusion Detection and Prevention Technique for
Misdirection Attack. The network parameters calculated by the
use of this technique shows considerable amount of improvement
in throughput while introducing small amount of delay.
Index Terms— Cluster Based Intrusion Detection and
Prevention Technique, End-to-end Delay, Misdirection Attack,
Misdirection Attack Detection and Prevention Algorithm,
Throughput.
I. INTRODUCTION
W IRELESS sensor nodes are low power electronic
devices, deployed in remote areas where power
resources are limited. The demand of Wireless
Sensor Networks (WSN) has extended to too many real world
applications such as habitat monitoring, environment
monitoring, military surveillance, etc. Sometimes the through
Roshan Singh Sachan, student is with the Department of Computer
Science and Engineering, Graphic Era University, Dehradun, 248002, India
(Phone: +918957272790; (e-mail: rsachan28@gmail.com).
Mohammad Wazid student is with the Department of Computer Science
and Engineering, Graphic Era University, Dehradun, 248002, India
Avita Katal student is with the Department of Computer Science and
Engineering, Graphic Era University, Dehradun, 248002, India
D P Singh and R H Goudar is with the Department of Computer Science
and Engineering, Graphic Era University, Dehradun, 248002, India
978-1-4673-4866-9/13/$31.00 ©2013 IEEE
795
insecure medium. Thus, WSN can be easily attacked by
Denial-of-Service (DoS) attacks, which cause information loss
along with large energy expenditure. Therefore, securing the
links is important in designing a sensor network. In
misdirection attack the attacker routes the packet from its
children to other distant nodes, but not necessarily to its
legitimate parent. This produces long delay in packet delivery
and decreases the throughput of the network. So detection and
prevention is required in misdirection prone WSN. Here we
have proposed a Cluster Based Intrusion Detection and
Prevention Technique for Misdirection Attack, which is
capable of detecting and defending against misdirection attack.
The produced results shows considerable amount of increment
in the throughput of the network with acceptable amount of
increment in delay. Section II includes the literature review
about different kinds of work done by the various authors
related to misdirection (DoS) attack. The novelty of the
proposed idea is discussed in section III. In section IV a brief
introduction to Misdirection attack is given. Section V
describes cluster based intrusion detection and prevention
technique for misdirection attack. The simulation scenario for
misdirection attack is discussed in section VI. In section VII
the key findings of this work are present followed by
conclusion, future work and applications in section VIII.
II. LITERATURE REVIEW
In paper [1] authors design a novel message observation
mechanism (MoM) to detect and defense the DoS attack.
Based on the spatiotemporal correlation, MoM utilizes the
similarity function to identify the content attack as well as the
frequency attack. The MoM adopts rekey and reroute
countermeasures to isolate the malicious node. The security
analysis shows that their solution not only detects and defenses
the DoS attack but also can reduce the energy consumption. In
[2] the advances in the area of the micro-electro-mechanical
systems (MEMS) technology provides to send the data without
any physical connection from Wireless Sensor Node (WSN) to
the Base Station, this data may be modified by an attacker or
dropped, if no security is provided. A Survey of various
secured routing protocols along with various authentication
methods which use minimum resources to resist against
various attacks is done. In paper [3] a new method is invented
for discovering the misdirection attack. Produced results show
that the misdirection attack causes consumption of network
resources. The advantage of this work is to show the
importance of reducing energy and time consumption as they
are very significant in data transmission and reducing them
makes transmission process more efficient and reliable. In
paper [4] two modifications to the Lightweight Medium
Access Control (LMAC) protocol are proposed and evaluated.
The first is Data Packet Separation Slot Size Randomization
(DSSSR); the second is Round Robin (RR) slot size
assignment. The paper shows that (DS-SSR) can increase the
WSN resistance against the Energy efficient denial of service
link layer jamming attacks. The paper also shows that
employing RR slightly eliminates the negative impact on the
network throughput when using countermeasures against
energy efficient jamming. Two measures are used to evaluate
the resistance of the proposed protocol against the attack:
lifetime advantage and censorship rate. Experimental results
show that about 8 % reduction of the attacker lifetime
advantage can be achieved with DS-SSR LMAC compared to
other countermeasure. Furthermore, results show that the
censorship rate of the proposed protocol was similar to the
other countermeasures. In this paper [5] the authors focused on
security of WSN with the conclusion that the security of
significant systems should be continually reassessed to take
new detections into account. The level of security needed from
the application should also be marked when preferring
hardware. Paper [6] proposes TARF, a trust-aware routing
framework for WSNs, to secure multi-hop routing in WSNs
against intruders exploiting the replay of routing information.
With the idea of trust management, TARF enables a node to
keep track of the trust worthiness of its neighbors and thus to
select a reliable route. Not only does TARF circumvent those
malicious nodes misusing other nodes’ identities to misdirect
network traffic, it also accomplishes efficient energy usage. In
[7] the challenges and potential solutions to achieve
dependability of WSNs taking into account accidental failures
as well as intentional attacks are discussed. In this article the
authors have identified scientific challenges with respect to
security and dependability that need to be solved to make
WSNs a ready to use technology for CIP. In order to cope with
those challenges they apply a vertical approach, where they do
not focus on a single layer or subsystem, but consider the
whole system in all its facets. In this paper [8], a new approach
to achieve confidentiality in multi-hop code dissemination is
presented. They integrate confidentiality and DoS-attack-
resistance in a multi-hop code dissemination protocol. The
approach is based on Deluge, an open source, state-of-the-art
code dissemination protocol for WSNs. In addition, they
provide a performance evaluation in their scheme, compared
with the original Deluge and the existing secure Deluge. This
paper [9] presents new method based on services and Markov
Chain. It distinguishes services, analyzes influence of services
under DoS attacks and gets the formula of survivability.
Probability of states is obtained by markov chain and then the
quantities survivability is obtained in different states.
Simulations prove that DoS attacks obviously decrease the
survivability of WSN but adding node energy and improving
796
node density in certain range can reduce the influence of DoS
attacks. Simulations prove DoS attacks decrease the
survivability of WSN, adding node energy and properly
improving node density can reduce the influence of DoS
attacks. A new method has been invented in paper [10] for
discovering the misdirection attack. The advantage of this
work is to show the importance of reducing energy and time
consumption. As these two factors are very significant in data
transmission and reducing them make transmission process
more efficient and reliable. Paper [13] depicts that many
threats can be defeated by using existing encryption and
authentication mechanisms and other techniques can alert
network administrators of ongoing attacks or trigger
techniques to conserve energy on affected devices. In paper
[14] to overcome vulnerabilities for large-scale static WSNs, a
location-aware end-to-end security framework in which secret
keys are bound to geographic locations and each node stores a
few keys based on its own location. This location-aware
property effectively limits the impact of compromised nodes
only to their vicinity without affecting end-to-end data
security. The proposed multifunctional key management
framework assures both node-to-sink and node-to-node
authentication along the report forwarding routes. Moreover,
the proposed data delivery approach guarantees efficient en-
route bogus data filtering and is highly robust against DoS
attacks. The evaluation demonstrates that the proposed design
is highly resilient against an increasing number of
compromised nodes and effective in energy savings. The
design is evaluated through extensive analysis, which
demonstrates its high resilience against an increasing number
of compromised nodes and effectiveness in energy savings that
is, achieving 85 percent or more energy savings in contrast to
the case without using our design when appropriate parameters
are chosen. This paper [15] presents an overview of the
various vulnerabilities in WSN at network layer, and the
defensive measures that can be taken against these threats.
Various security protocols for sensor networks have been
evaluated against these vulnerabilities. Most of the attacks on
some well known and widely used security protocols have
been discussed, along with the possible countermeasures for
these attacks. Some new emerging protocols have also been
discussed in this context. Suitability of implementation of
cryptographic techniques at the network layer has also been
investigated for these severely resource constrained networks.
Although most of the threats have been addressed by
researchers much deeply, but a need exists to combine all these
separate efforts to develop a unique light weight and powerful
protocol to address the issues cumulatively. Paper [16] checks
the qualitative analysis of various proactive and reactive
protocols which compromise with the internal attacks (i. e.
misdirection attack) and whether one type of protocol offers
better resistance to various attacks than the other. In this [17] a
novel RSA-based framework for combating DoS attacks in
WSN by ensuring that the malicious party will exhaust its
resources prior to exhausting those of its counterparts is
proposed. Under the proposed approach, the mathematical
operations performed by the malicious party require two or
three orders of magnitude more resources than those required
by the attacked party. The authors also present three
methodologies for establishing an ephemeral key, in which the
proposed DoS mitigation mechanism is an embedded
component. Implementation results on the Intel Mote 2
platform substantiate the clear advantages of the proposed
method the concept developed can be applied to a wide range
of additional security services that are currently not offered in
WSN environments. This paper [18] addresses an especially a
very malignant type of of DoS attack, called PDoS (Path-based
Denial of Service) in which an adversary overwhelms sensor
nodes a long distance away by flooding the node with replayed
Fig. 2. Flow Packets when R1 becomes Attacker
packets or injected spurious packets. A solution using one-way
hash chains to protect end-to-end communications in WSNs Figure 2 shows the flow of packets; they are misdirected to
against PDoS attacks is proposed. The solution provided is node R2 by the malicious node R1. In this paper we have
lightweight, tolerates burst packet losses, and can easily be simulated this attack.
implemented in modern WSNs. 2) Packets forwarded to a node far away from the
destination: This kind of misdirection attack is very harmful
III. PROBLEM DEFINITION AND NOVELITY because all packets are forwarded to a node far away,
preventing them to reach the destination so packets will not
Misdirection Attack is one of the alarming attacks in WSN reach destination. Due to the attack the delay becomes infinite
as it reduces the throughput of the network along with the and further results in zero throughput.
introduction of high end to end delay. In this paper a novel
technique for intrusion detection and prevention of V. INTRUSION DETECTION AND PREVENTION
misdirection attack is proposed. It is a cluster based approach TECHNIQUES FOR MISDIRECTION ATTACK
for the detection and prevention of misdirection attack.
Proposed technique is very effective in the detection and A. Cluster Based Intrusion Detection and Prevention
prevention of this attack. We have also measured the impact of Technique
this technique on the performance of network (i. e. End-to-end
delay and Throughput). In misdirection attack malicious nodes misdirect packets to
other nodes but not to the intended recipient. As the malicious
IV. STUDY OF MISDIRECTION ATTACK intermediate node (routing node) begins misdirecting packets
due to this packets reach destination but not from the actual
It is the most popular Denial of Service Attack. This attack
path but from some other path producing long delay in the
can be performed in different ways. A malicious node could
network due to which throughput is also decreased.
deny a valid route to a particular node thereby denying service
The Cluster based intrusion detection and prevention
to the destination.
technique introduced in this paper detects the misdirection
attacker node and also prevents this attack from occurrence.
A. Types of Misdirection attack
In the proposed technique we build clusters from mobile
It can be performed in two ways:
nodes. Cluster contains nodes which are in the communication
range of each other. A cluster head is elected by these nodes
1) Packets forwarded to a node near to the destination: This
for each cluster. Election of a cluster head requires two things
kind of misdirection attack is less intense, because packets
to be considered.
reach to the destination but from a different route which
x Fairness: Any node can become cluster head meaning
further produces long delay, thus decreasing throughput of
that the probability of every node to be a cluster head
network (bit transfer per second).
should be equal.
x Efficiency: There should be some method that can
select a node from the cluster periodically which has
high efficiency (i.e. battery backup).

When cluster is formed, cluster head is elected and it

becomes the responsibility of the cluster head to detect the
intruder node in that cluster. Source node maintains a FIFO
buffer corresponding to each packet. This buffer contains entry
Fig. 1. Normal flow of Packets
of each sent packet with time stamp value corresponding to
Figure 1 shows normal flow of packets. each packet sequence number. Source node also shares this
buffer to the cluster head. Cluster head compares all sequence
numbers of packets stored in its buffer to the sequence
numbers of packets stored in buffer of all intermediate nodes

797
with stamp value. If packet mismatch or empty entry is found
in the buffer at a particular node, then the previous node will
be omitted. The detection process again starts right from the
beginning. It again searches for another optimum route for the
secure communication. Thus any misdirection attack is easily
detected and prevented with the proposed technique.
Fig. 3. Normal Flow of Datagrams
Figure 3 shows the normal flow of traffic in the network
without any intruder. At each node there is a buffer having the
timestamp (time_stamp) value and corresponding packet
sequence number (pack_sqno). Packets are transferred from
node N1 to node N5 via N2 and N3.
Fig. 4. Buffer of node N1
Figure 4 shows buffer entries i.e. time stamp values and packet
sequence numbers.
Fig. 5. Flow of packets when sensor node SN2 becomes misdirection attacker
Figure 5 shows flow of traffic when sensor node SN2 becomes
misdirection attacker. Initially all the traffic is transmitted via
SN2 and SN3 to SN5 (destination node). When sensor node
SN2 becomes attacker it misdirects packets to sensor node
SN6. Now packets are coming through sensor node SN6-SN7-
798
SN4 to SN5 (destination). Thus delay will be increased in the
presence of misdirection attacker node (i.e. SN2).
Fig. 6. Intrusion Detection using Proposed Technique
Figure 6 shows that node SN2 becomes intruder. In detection
and prevention technique cluster head compares the buffer
which was sent by source node to the buffer maintained at all
intermediate nodes. Time stamp value at source and also at
cluster head is t, suppose packet takes t1 units of time from in
reaching from one node to other node i.e. 2 seconds. So value
of time stamp in Buff_N3 will become t’=> 1+2=3 seconds
and so on.
Fig. 7. Mismatch detection in buffers when some packets are misdirected
Figure 7 shows the mismatch in packet sequence numbers
(pack_sqno) because of the misdirection attack node by
intruder sensor node SN2. Packets 1, 2, 3 are misdirected here.
This is detected by cluster head because of the buffer which
was sent by the source node.
Fig. 8. Mismatch detection in buffers when all packets are misdirected
Figure 8 shows the mismatch in packet sequence numbers
(pack_sqno) because of the misdirection attack done by
intruder sensor node SN2. In the pack_sqno filed we get empty
values because all packets are misdirected by intruder sensor
node SN2 to sensor node SN6 so on. This is detected by
cluster head because of the buffer which was sent by the
source node.
 Complexity Analysis:

=
Fig. 9. Normal flow of packets after the deployment of proposed Technique
= 1+2 +3 + ----------- + n= n (n+1)/2
Figure 9 shows a new routing path, formed after intrusion If the value of the timer is t
detection process by omitting the intruder node (SN2) and a = t (n (n+1)/2)
normal flow of traffic is again achieved in the network. Now If the number of nodes are x
traffic is transmitted via new formed path (SN1-SN3-SN4- = x t (n (n+1)/2)
SN5) in place of (SN1-SN2-SN3-SN5). x and t are constant values.
Therefore the complexity is Ө (n2)
B. Misdirection Attack Detection and Prevention Algorithm

For detection and prevention purpose we have proposed C. Flowchart of the proposed Method
following algorithm:

Elect a cluster head using any election algorithm.

RETURN: Discover an optimum route from source to
destination.
Set timer (t) at CH.
Start packet transmission from source to destination on the
selected route
Source maintains a buffer corresponding to each packet
// t is the time when packet was sent by source
for iÅ1 to n do
assign number to each packet , pack[t][i];
Source shares this buffer to CH
Maintains a receiving buffer at each intermediate node at
time= t’
// suppose packet takes t1 units of time from in reaching
from one node to other node.
t’=t + t1;
for iÅ 1 to n do
assign number to each packet , recvbuffpack [t’][i];
Cluster Head periodically compares its buffer to receiving
buffer at each intermediate node.
for each node[x] do
for iÅ 1 to n do
for jÅ 1 to i do
compare pack[t][i] and recvbuffpack [t’][j];
//two possibilities are there all the packets are misdirected
to some other node then in that case the receiving buffer of
node [x] is empty or some packets are misdirected then it
has some entries like packet 1, 2, 3 are misdirected but
packet 4 is directed towards intended recipient.
if (pack[t][i] ≠recvbuffpack [t’][j] or
recvbuffpack[t’][]) then do

Fig. 10. Diagrammatic representation proposed method

799
 TABLE I
COMMON PARAMETERS USED IN SIMULATION
VI. SIMULATION SCENARIO OF MISDIRECTION
ATTACK IN WSN Parameter Value

A. Simulation Scenarios Platform Windows XP SP2

To verify this work we simulate a WSN under misdirection Simulator Opnet modeler 14.5
attack. We have used the following two simulation scenarios in Area 500x500 met (Fix)
this paper:
14 Sensor Nodes
Network Size 2 Routers
1 Coordinator
Topologies Tree
Simulation Time 30 Minutes
Packet Size (bits) Constant (1024)

3) Implementations of Misdirection Attack: Under normal

Fig. 11. WSN with Normal Flow
Figure 11 shows normal flow of packets. In this scenario we
have considered 14 sensor nodes (i.e. S1, S2 etc), two routers
(i.e. R1, R2). Under normal flow sensor nodes S1, S2, --- S10
are reporting to router R1 and S11, S12, --- S14 are reporting
to router R2. Router R1 and R2 both reporting to the
coordinator.
flow all sensor nodes are sending packets to router R1 which
further sends them to coordinator. Router R2 is also reporting
to coordinator. When misdirection attack has happened router
R1 becomes attacker and starts misdirecting coming packets
towards Router R2. Router R2 is also in the communication
range of coordinator so all packets ultimately reach at
coordinator but with some delay and reducing throughput. We
have simulated this work under tree topology.
C. Results
In simulation we have taken following statistics of the
network: End-to-end Delay (msec), Throughput (bps). The
complexity of proposed algorithm is in c*n2. Number of sent
packets per second are 1.84. Therefore n2 is (1.84)2= 3.38 .The
simulation runs for 1800 seconds. So total delay produced by
this algorithm will be 3.38*1800 / (1000) 2 = 0.0061 units.
Total delay under proposed method would be equal to delay
produced under normal flow and delay under proposed
algorithm i.e. 14.02 + 0.0061= 14.03 msec.
Total decrement in throughput would be 12805.70 * 0.0061=
77.94 bps. So throughput in the presence of proposed method
= 12805.70 - 77.94 = 12727.76 bps.

TABLE III
END-TO-END DELAY & THROUGHPUT

Normal Under Misdirection Under Proposed

Flow Attack Method
End-to-end 14.02 14.11 14.03
Delay (msec)
Throughput 12805.70 11478.29 12727.76
Fig. 12. WSN under Misdirection Attack (bps)

Figure 12 depicts misdirection attack. In this scenario we have

again considered 14 sensor nodes (i.e. S1, S2 etc), two routers
(i.e. R1, R2). Under misdirection attack the packets which are
coming to router R1 are misdirected to router R2. The
malicious node R2 is shown in red label.
B. Experiment Design Parameters
1) Common Parameters:

800

Fig. 13. End-to-end Delay (msec) Fig. 14. Throughput (bps)
Figure 13 shows end-to-end delay (msec) with normal flow,
under attack and under proposed algorithm. Figure 14 shows
throughput (bps) with normal flow, under attack and under
proposed algorithm.
VII. KEY FINDINGS
In this paper we have measured the impact of the intrusion
detection and prevention techniques on the performance of
WSN under the presence of misdirection attack. Some of the
observations are:
x The End-to-end delay for normal flow is 14.02 msec,
14.11 msec under misdirection attack which further
improves to 14.03 msec when proposed method is
applied (Refer Table III). The Throughput of the
network for normal flow is 12805.70 bps, 11478.29
bps under misdirection attack which further improves
to 12727.76 bps when proposed method is applied
(Refer Table III).
VIII. CONCLUSION
x The presence of misdirection attack affects entire
performance of network especially throughput and
End to end delay. The proposed cluster based
intrusion detection and prevention technique is very
effective to detect and prevent misdirection attack.
Throughput has increased considerably by proposed
method but it has introduced some delay.
x In future this can be extended with some other
topology; we can also vary node density etc.
x We can apply proposed detection and prevention
technique in misdirection attack prone wireless sensor
networks which gives us secured network with
improved performance.
REFERENCES
[1] Yi-ying ZHANG, Xiang-zhen LI, Yuan-an LIU, “The detection
and defence of DoS attack for wireless sensor network” , Elsevier
Journal of China Universities of Posts and Telecommunications,
Volume 19, Supplement 2, October 2012, Pages 52-56.
[2] Shashikala , C. Kavitha, “A Survey on Secured Routing Protocols
for Wireless Sensor Network”, IEEE ICCCNT'12, Coimbatore,
India, 26-28th July 2012.
[3] Syed Omar, Tariq Saleh, Al-Khalidy, “Design an Algorithm To
Discover The MisdirectionAttack For Increasing The Life Time in
ComputerNetwork”, International Journal of Computer Science
and Information Security (IJCSIS), vol. 9, No. 1, December 2011.
[4] R. Ahmed Mahmood , H. Hussein Aly , N. Mohamed El-Derini,
“Defending Against Energy Efficient Link Layer Jamming Denial
of Service Attack in Wireless Sensor Networks” , IEEE AICCSA
2011.
[5] Hero Modares, Rosli Salleh, Amirhossein Moravejosharieh,
“Overview of Security Issues in Wireless Sensor Networks”, IEEE
3rd International Conference on Computational Intelligence,
Modelling & Simulation, 2011.
801
[6] Guoxing Zhan, Weisong Shil, Julia Deng, “TARF: A Trust-Aware
Routing Framework for Wireless Sensor Networks”, EWSN 2010,
LNCS 5970, pp. 65–80, 2010.
[7] N Levente Buttya ,Dennis Gessner, Alban Hessler, Peter
Langendoerfer, , “Application of Wireless Sensor Networks in
Critical Infrastructure Protection: Challenges and Design
Options”, IEEE Transactions on Wireless Communications
October 2010.
[8] Hailun Tan,Diethelm Ostry, John Zic,Sanjay Jha, “A Confidential
and DoS-Resistant Multi-hop Code Dissemination Protocol for
Wireless Sensor Networks”, ACM WiSec’09, Zurich, Switzerland,
March 16–18, 2009.
[9] Jiang Zhongqiu,Yan Shu,Wang Liangmin, “Survivability
Evaluation of Cluster-Based Wireless Sensor Network under DoS
Attacks” IEEE 5th International Conference on Wireless
Communications, Networking and Mobile Computing (WiCom
'09), 2009.
[10] M. Y. Abdullah, Hua Gui Wei, N. Alsharabi, “Wireless sensor
networks misdirection attacker challenges and solutions”, IEEE
International Conference on Information and Automation, 2008.
[11] Somanath Tripathy, Sukumar Nandi, “Defense against outside
attacks in wireless sensor networks”, Elsevier Computer
Communications, Volume 31, Pages 818-826 Issue 4, 5 March
2008.
[12] An Liu, Young-Hyun Oh, Peng Ning, “Secure and DoS-Resistant
Code Dissemination in Wireless Sensor Networks Using Seluge”,
ACM International Conference on Information Processing in
Sensor Networks 2008.
[13] David R. Raymond, Scott F. Midkiff, “Denial-of-Service in
Wireless Sensor Networks: Attacks and Defenses”, Published by
the IEEE Pervasive Computing magazine no 1536-1268/08/ 2008.
[14] Kui Ren , Wenjing Lou, Yanchao Zhang “LEDS: Providing
Location-Aware End-to-End Data Security in Wireless Sensor
Networks”, IEEE Transactions on Mobile Computing, vol. 7, No.
5, May 2008.
[15] Asif Habib, “Sensor Network Security Issues at Network Layer”,
IEEE 2nd International Conference on Advances in Space
Technologies, (ICAST’ 08) 2008.
[16] Po Wah Yau, Shenglan Hu, Chris J. Mithell, “Malicious attacks
on ad hoc network routing protocols”, International Journal of
Computer Research Vol 15 Issue 1, 2007.
[17] Arazil, H. Qi1, D. Rose “ A Public Key Cryptographic Method for
Denial of Service Mitigation in Wireless Sensor Networks”, 4th
Annual IEEE Communications Society Conference on Sensor,
Mesh and Ad Hoc Communications and Networks, ( SECON '07)
2007.
[18] Jing Deng, Richard Han, and Shivakant Mishra, “Defending
against Pathbased DoS Attacks in Wireless Sensor Networks”,
ACM SASN’05, November 7, 2005.