Wireless Sensor Network, 2013, 5, 33-40
http://dx.doi.org/10.4236/wsn.2013.53005 Published Online March 2013 (http://www.scirp.org/journal/wsn)
A Secure Routing Method for Detecting False Reports and
Wormhole Attacks in Wireless Sensor Networks*
Hyeon Myeong Choi, Su Man Nam, Tae Ho Cho
College of Information and Communication Engineering, Sungkyunkwan University, Suwon, South Korea
Received May 16, 2012; revised June 14, 2012; accepted June 25, 2012
ABSTRACT
Wireless sensor networks (WSNs) consist of a large number of sensor nodes that monitor the environment and a few
base stations that collect the sensor readings. Individual sensor nodes are subject to compromised security because they
may be deployed in hostile environments and each sensor node communicates wirelessly. An adversary can inject false
reports into the networks via compromised nodes. Furthermore, an adversary can create a wormhole by directly linking
two compromised nodes or using out-of-band channels. If these two kinds of attacks occur simultaneously in a network,
existing methods cannot defend against them adequately. We thus propose a secure routing method for detecting false
report injections and wormhole attacks in wireless sensor networks. The proposed method uses ACK messages for de-
tecting wormholes and is based on a statistical en-route filtering (SEF) scheme for detecting false reports. Simulation
results show that the proposed method reduces energy consumption by up to 20% and provide greater network security.
Keywords: Wireless Sensor Network; Security; Statistical En-Route; Wormhole; Secure Routing
1. Introduction
Recent advances in micro electro mechanical systems
and digital electronic and wireless communication tech-
nologies have enabled the development of low-cost,
multi-functional sensor nodes [1]. Wireless sensor net-
works (WSNs) have long been used in many types of
computing systems. A WSN consists of a large number
of sensor nodes that monitor the environment and one or
more base stations that collect the sensor readings [2]. In
many applications such as military surveillance, sensor
nodes are deployed in open, large-scale, and even hostile
environments and potential issues range from accidental
node failure to intentional tampering. Due to their rela-
tively small sizes and unattended operations, sensor nodes
are at high risk of being physically captured and having
their security compromised [3]. Additionally, the power
of the sensor nodes is limited and non-replaceable [4-6].
The security and energy efficiency of sensor nodes are
thus extremely important in WSNs.
If sensor nodes are physically captured and compro-
mised, security information such as network keys can be
revealed to the adversary. The adversaries can then inject
false reports into sensor network via the compromised
nodes. These injected false reports can not only result in
false alarms but also in quick usage of the limited
amount of energy in the sensor nodes [7]. Several re-
searchers have proposed mechanisms to combat attack by
injection of false reports [7-14]. The statistical enrouting
filtering scheme (SEF) was proposed by Fan Ye et al. [7]
to detect and drop injected false reports during the for-
warding process.
Another type of wireless sensor network attack is a
wormhole attack, which is made up of two adversaries
and a wormhole tunnel. The two adversaries communi-
cate with each other through the wormhole tunnel, which
is a direct and dedicated channel using a wired link or
additional RF transceivers on an out-of-bound channel
available only to the attacker. A wormhole attack can
alter or drop messages as well as eavesdrop. To combat
this type of attack, a few different countermeasures have
been proposed [2,4-6,9-13,15-18]; a simple lightweight
protocol called LITEWORP is one of these. LITEWORP
uses a secure two-hop neighbor discovery and local
monitoring of control traffic to detect nodes involved in a
wormhole attack [17].
In this paper, we propose a secure routing method for
detecting multiple attacks in wireless sensor networks,
particularly false report injections and wormholes. Pre-
vious research has focused on single attacks (e.g., either
false report injections or wormholes); there are currently
no reports of research in cases in which the two types of
attacks occur in a network at the same time. To defend
*This research was supported by Basic Science Research Program
through the National Research Foundation of Korea (NRF) funded by
the Ministry of Education, Science and Technology (No. 2012-
0002475).
C
opyright © 2013 SciRes. WSN

H. M. CHOI ET AL. 34
against multiple attacks, it is possible to simply imple-
ment two countermeasures (i.e., SEF and LITEWORP)
together; however, because each individual countermea-
sure does not consider the use of the other, this approach
can result in wasteful energy consumption. In the pro-
posed method, we consider both false report injections
and wormhole attacks, leading to prevention of addi-
tional energy consumption. Simulation results show that
the proposed method can save up to 20% of total energy
consumption compared to the simple combination of the
SEF and LITEWORP, while simultaneously providing
greater security.
The remainder of this paper is organized as follows:
Section 2 briefly describes false report injections and
wormhole attacks. Section 3 presents the problem state-
ment, followed by a detailed description of the proposed
method in Section 4. In Section 5, the performance of the
proposed method is shown via simulation and discussed.
Finally, Section 6 presents the conclusions of our study.
2. Background
2.1. Attack Models and Related Works
Figure 1 shows a schematic of a false report injection
attack. A compromised node can inject false reports into
the network, which both wastes the limited energy of the
many nodes that deliver them to the base station (BS)
and lead to false alarms. Several reports in the literature
have proposed methods to combat false report injection
attacks [7-14], including one by Fan Ye et al. [7], which
used statistical en-route filtering (SEF) as a means to
detect and drop injected false reports during the for-
warding process. SEF carefully limits the amount of se-
curity information assigned to each node to prevent any
Figure 1. False reports injection attack model.
single compromised node from disrupting the entire sys-
tem. It relies on the collective decisions of multiple sen-
sors for false report detection.
Figure 2 shows a schematic of a wormhole attack. The
nodes X and Y are compromised and form a wormhole
tunnel connected via wired link or a powerful out-of-
bound RF channel. If node S wants to send a message to
the BS, the routing path that includes the wormhole link
has an advantage because its hop count is lower than that
of the normal routing path. Messages routed via a
wormhole node can also be dropped or altered; it is thus
important that messages are not routed through worm-
hole nodes. To combat these wormhole attacks, Hu et al.
[15] proposed geographical and temporal packet leashes
as techniques for detecting wormholes. For the geo-
graphical leash, the sender appends its location and the
sending time to the packet. Based on this information, the
receiving node computes an upper bound on the distance
to the sender. In the temporal leash, the sender appends
the sending time to the packet and the receiving node
computes a traveling distance for that packet using the
difference between the packet sending and receiving
times, assuming propagation at the speed of light. To
implement these two schemes, the geographical leashes
require that all nodes have a localization system such as
GPS, and the temporal leashes require accurate local
clocks and global time synchronization of all nodes in the
network. For instance, Lingxuan et al. [16] proposed use
of a directional antenna and ultrasonic signals to address
both of these issues. However, the schemes mentioned
above require that each sensor node be equipped with
special devices. In LITEWORP [17], the neighboring
nodes common between two nodes are chosen as guard
nodes. The guard node monitors all traffic from both
nodes and verifies that both nodes are free of malicious
behavior; via this kind of monitoring, guard nodes can
detect selective forwarding by a wormhole attack. Addi-
Figure 2. Wormhole attack model.

H. M. CHOI ET AL. 35
tionally, LIny special
2.2. Statistical En-Route Filtering (SEF)
F sche-
2.2.1. Key Assi gnment and Report Generation
t into n
su
ro
2.2.2. En-Route Filtering and BS Verification
h inter-
er of MACs in the reports. If t
re
that the key indices in the attached report
be
the node has a key that matches that of the report,
th
operations 1-3, the node sends it
to
detect false reports en-route, but ev
fa
or Wormhole
Attacks (LITEWORP)
tack alized hardware such as
and outgoing data packet of its neighbor
the guard node revokes
t
of a combined false report
ck in a wireless sensor net-
ay contain some duplication be-
TEWORP can operate without a
devices such as GPS. However, the guard nodes must
monitor all traffic between the two nodes, resulting in
increased processing overheads; as the guard nodes are
not special nodes, such overhead invariably shortens
their lifetime.
In this section, the main characteristics of the SE
me are described. Similar to the general en-route scheme,
the SEF also involves two primary phases.
The BS maintains a global key pool and divides i
partitions. Each partition has m keys, and each key has a
unique key index. Before the nodes are deployed, the
user randomly selects k keys from one partition. The
selected keys and the associated key indices are stored in
the node before being deployed to the sensing field.
When an event occurs in the sensing field, all r-
nodes. When a node sends a data packet to a receiv-
ing node, the guard nodes save the packet information
in a watch buffer. The information includes the
packet’s identification and type, source, destination,
and immediate sender and receiver. The guard nodes
expect that the receiving node will forward the packet
toward the base station unless the receiving node is
itself the base station. Each entry in the watch buffer
has a time threshold; the receiving node must send the
packet onwards before the time threshold expires. A
malicious activity counter is maintained by each
guard node. The malicious activity counter is incre-
mented for each neighbor node malicious event that is
detected by the guard node.
 Isolation: When the malicious activity counter of
node A crosses a threshold,
unding nodes detect the event and one of these nodes is
elected as the CoS (Center of Stimulus) node. All detect-
ing nodes generate a message authentication code (MAC)
via one of their keys and send it to the CoS node along
with the key index. The CoS node collects and classifies
MACs based on the key partitions. Then the CoS node
generates a report consisting of the event information, the
MACs chosen from the distinct partitions and the key
indices used to generate the MACs. The number of
MACs included in each report is exactly same for all
reports.
As reports are forwarded via multiple hops, eac
mediate forwarding node is able to verify them via the
following operations:
1) Check the numbhe
node A from its neighbor list, then sends alert mes-
sages to each neighbor node of node A indicating that
node A is a suspected malicious node. When a
neighbor node X of node A receives the alert, it stores
the identity of the guard node in an alert buffer asso-
ciated with A. When the number of alert messages
regarding node A is over the threshold for node X,
node X removes node A from its neighbor list. After
isolation, node X does not receive or send any packet
to a revoked node.
3. Problem Statemen
port has a different number of MACs, the node drops
the report.
2) Check
long to distinct partitions. If the report has more than
one key index from the same partition, the node drops the
report.
3) If
e node generates a MAC with that key. If generated
MAC and the corresponding MAC are different, the node
drops the report.
4) If the report passes
the next hop.
The SEF can en if cause they are not designed to work together. For ex-
ample, for sensor networks using both countermea-
sures, the SEF attaches message authentication codes
(MACs) to reports for detection of false reports, in-
lse reports arrive at the BS, the BS is able to verify
every MAC because it has all the keys. If there are any
mismatches, the BS discards the report.
2.3. Lightweight Countermeasure f
LITEWORP [17] is a countermeasure for wormhole at-
s that does not require speci
GPS. In the LITEWORP scheme, neighboring nodes
common between two nodes are chosen as their guard
nodes, which monitor the incoming and outgoing traffic
of their neighbors. LITEWORP is operated in two phases
as follows:
 Wormhole Detection: Guard nodes monitor every
incoming
Figure 3 shows a schematic
injection and wormhole atta
work. Generally, to combat this multiple attack, the net-
work can use two countermeasures to deal separate with
the false report injection attacks (i.e., SEF) and worm-
hole attacks (i.e., LITEWORP). These two countermea-
sures are not designed to work together, so a few prob-
lems occur as follows:
 Energy consumption: The two countermeasures, SEF
and LITEWORP, m

H. M. CHOI ET AL. 36
Figure 3. Schematic of two attacks occurring simultane-
ously in a sensor network.
ach report. LITEWORP moni-
tors all traffic in the network for wormhole attack de-

ports at an early

se reports without any other opera
the node two hops down the line. The
s.
4.
sensor network composed of a
BS. We also
multiple hops
per, we propose a secure routing method for
report injection and wormhole attacks
or networks. The proposed method is
l is di-
r to generate reports. The CoS node collects
creasing the size of e
tection, but the large reports generate high overhead
and wasteful energy consumption.
Security Level: The best way to combat false data
injection attacks is to detect false re
stage and drop it immediately. On the other hand, the
key countermeasure to combat wormhole attacks is to
find abnormal behavior in the network and drop
nodes with these abnormal behaviors. One behavior
considered abnormal is selective forwarding, the re-
fusal of malicious nodes to forward certain reports,
resulting in dropped messages not delivered to the
base station. Because these two countermeasures do
not communicate with each other, nodes enacting the
false report injection countermeasure that drops false
reports are considered by the wormhole countermea-
sure to be part of a wormhole.
To address these problems, we introduce improve-
ments as follows:
The false report detection method is based on SEF,
but SEF drops fal-
vided into n partitions in the BS. Each node stores k
keys from one of these partitions. Then the nodes are
deployed in the sensing field. After nodes are de-
ployed, each node sends hello messages to recognize
their neighbor nodes. The BS sends routing messages
or operates other routing method to establish routing
paths.
 En-route filtering: When an event occurs in the sens-
ing field, the surrounding sensing nodes compete with
each othe
tions such as a message to other nodes notifying the
dropping of a false report. The proposed method
sends the key index to the base station when false re-
ports are dropped; this key index message can be used
to notify the base station and other nodes of a dropped
false report.
 All nodes that send reports must receive an ACK
message from
M
ACK messages are transmitted in a different way
from the reports and include the node IDs on the
routing path. A detailed description of these ACK
messages is presented in the following section.
The analysis and simulation results described in Sec-
tion 5 show the effectiveness of these improvement
Proposed Method
4.1. System Model and Assumptions
We consider a large scale
large number of small sensor nodes and a
assume that the reports are forwarded via
toward the BS. The sensor nodes are not moved after
deployment. We also assume that the energy at the BS is
unlimited. The sensor nodes are small and operated indi-
vidually, so attacker can compromise a sensor node and
inject false reports into the network with it. Moreover,
the attacker can make a wormhole using a laptop or an-
other similar device. We also assume that false report
injection attacks occur with wormhole attacks in the
network.
4.2. Overview
In this pa
detection of false
in wireless sens
more efficient than one that uses two different methods
simultaneously because the two different defense me-
thods are not designed to work with each other. Addi-
tionally, it solves the problems that occur in previous
solutions and provides efficient energy consumption. The
proposed method is divided into three phases: Initializa-
tion, en-route filtering, and wormhole detection.
 Initialization: This phase is operating before node
deployment and immediately after node deployment.
Before node deployment, the global key poo
ACs from collaborating sensing nodes, attaches
them to the report and forwards it towards the BS.
Each intermediate forwarding node can verify the re-
ports and drop false reports. When a false report is
dropped, the intermediate node sends a drop message
to the next node. If the intermediate node does not
send a drop message, the previous node cannot re-
ceive an ACK message, causing the previous node to
regard the intermediate node as a wormhole link. To
prevent this problem, the intermediate nodes send
drop messages to the next node when dropping false
reports.
 Wormhole detection: All nodes sending reports wait
for an ACK message. If nodes do not receive the

H. M. CHOI ET AL. 37
ACK messages, the next node is wormhole link. The
des are deployed, key assignment is based
he global key pool is divided into n
itions. Each partition has m keys,
he sensing field, all surround-
one of these nodes is
s node. The CoS node
it for ACK messages after
K messages do not arrive
de is regarded as a worm-
CK messages. The ACK messages are used
to
ACK messages must be transmitted between nodes
separated by two hops, but cannot be transmitted via
the path that the original report is sent on. Since the
ACK messages must be sent via other path, the time
to live (TTL) is important. The TTL is the maximum
number of hops used to transmit the ACK messages.
If the ACK messages cannot be delivered to the pre-
vious node within the TTL hop limit, a wormhole is
detected.
Overall, the focus of this method is on solving the
conflict between the false report and wormhole detection
mechanisms.
4.3. Initialization
Before any no
on the SEF scheme. T
non-overlapping part
and each key has a corresponding key index. The user
randomly selects one of the n partitions, and randomly
chooses k keys from it. The chosen key and key indices
are stored in the node. After all nodes store keys and key
indices, the nodes are deployed into the sensing field.
After deployment, all nodes immediately send hello
messages to their neighbor nodes. All neighbor nodes
replying to the hello message with their partition infor-
mation. This partition information is used to determine
alternative routing paths.
4.4. False Report Detection
When an event occurs in t
ing nodes detect the event and
elected as the center of stimulu
cooperates with the surrounding nodes to generate a re-
port. As a report is forwarded multiple hops, each inter-
mediate forwarding node can verify it via the SEF
scheme. If a node drops a false report, it sends a drop
message to the next node. Figure 4 shows that a com-
promised node sends a false report, but node B detects
this false report and drops it. If node B does not notify
the next node C of the false report drop event, node C
never sends an ACK message to node A, leading node A
to regard node B as a wormhole link. To prevent this
problem, the proposed method sends a drop message
after nodes drop false reports.
4.5. Wormhole Detection
Nodes that transmit reports wa
sending reports, but if the AC
until after time t, the next no
hole. Therefore the sending node eliminates the next
node in the routing path, and then retransmits the reports
to another node. Figure 5 shows the report transmission
process. The CoS node sends a report to the BS, and each
intermediate node sends an ACK message to detect
wormholes.
All nodes that transmit reports must wait for ACK
messages. It means all nodes that receive reports have to
reply with A
detect wormhole links. Figure 6 shows an example of
replying with ACK messages. When node A sends a re-
port to node B and node B sends it on to node C, node C
replies with an ACK message to a neighbor node common
to nodes B and C. Furthermore, nodes receiving ACK
messages (node X) also send ACK messages to neighbor
nodes common to node B and itself. This operation is
Figure 4. False report detection process.
Figure 5. Overview of the message transmission process.
Figure 6. ACK message transmission.

H. M. CHOI ET AL. 38
repeated until the ACK message is delivered to node A.
Figure 7 shows a report transmitted to node B from a
node A via a wormhole tunnel. Since node A sends the
reports, it waits for an ACK message, which it cannot
receive. The node B sends an ACK message to node A,
but the ACK message has a maximum hop limit called
Time-to-Live (TTL). If no limit was set, the ACK mes-
sages would float throughout the network, wasting limi-
ted node energy. Figure 7 shows that node B sends an
ACK message to node A with a limited TTL value, but
the ACK message cannot be delivered to node A. If the
TTL value is too large, it may be delivered to node A
even when reports are transmitted via a wormhole. If the
TTL value is too small, it may not be delivered to node A
even wheormhole.
ba
8(a) and the probability of
false alarms (assuming no false reports in the network)
b) herein, we calculate the
n reports are not transmitted via w
Therefore the TTL value must be carefully determined
sed on network statements.
5. Simulation Results
To show the effectiveness of the proposed method, we
compare the proposed method with the combined
LITEWORP and SEF schemes via simulation. Each node
consumes 16.25 μJ and 12.5 μJ to transmit and receive a
byte, respectively. The size of an original report and a
MAC is 24 byte and 1 byte, respectively. Each message
includes 5 MACs for the SEF. There is a global key pool
of 1000 keys, the number of partitions is 25 and each
node possesses 30 keys.
A probability comparison for wormhole detection be-
tween the LITEWORP+SEF method and the proposed
method is presented in Figure
are presented in Figure 8(
probability of wormhole detection based on the percent-
age of isolated wormhole nodes and the probability of
false alarms based on isolated normal nodes. The pro-
posed method shows a higher wormhole detection per-
formance than the LITEWORP+SEF method when the
number of neighbor nodes is less than 7 or greater than
21; however, the proposed method shows a lower false
alarm performance (higher probability) than the LITE
WORP+SEF method. Because the proposed method uses
ACK messages for wormhole detection, the ACK mes-
Figure 7. Wormhole detection.
sages cannot be delivered to the previous node if the
number of neighbor nodes is low.
Figure 9 shows the probability of false alarms versus
(a) Probability of wormhole detection
(b) Probability of false alarm
Figure 8. Probability of detecting wormholes and false
alarms.
Figure 9. Probability of false alarm.

H. M. CHOI ET AL. 39
Figure 10. Energy consumption for different false traffic ratios (FTR).
the ratio of false reports when TTL = 3 and the number
of neighbor nodes is 9. The proposed method has th
same probability of false alarms regardless of the false
traffic ratio. The LITEWORP+SEF method has a high
probability of false alarms when the false traffic ratio is
high because SEF detects false reports and drops them
without notifying neighbor nodes. LITEWORP monitors
this drop event and regards it as malicious behavior,
hereby increasing the probability of a false alarm.
Figure 10 shows the energy consumption of the pro-
posed method and the LITEWORP+SEF method. The
simulation results show that the proposed method con-
sumes less energy than the
every false traffic ratio (
e network. The LITEWORP+SEF me
t of energy. The proposed method
ms using ACK messages. The simula-
tion results show that the proposed method decreases
both the probability of false alarms and energy consump-
tion.
REFERENCES
[1] I. F. Akyildiz, W. Su, Y. Sankarasubramaniam and E.
Cayirci, “A Survey on Sensor Networks,” IEEE Commu-
nications Magazine, Vol. 40, No. 8, 2002, pp. 102-114.
doi:10.1109/MCOM.2002.1024422
e
[2] L. Buttyan, et al., “Statistical Wormhole Detection in
Sensor Networks,” Lecture Notes in Computer Science
Vol. 3813, 2005, pp. 128-141. doi:10.1007/11601494_11
,
LITEWORP+SEF method for
FTR) shown. The proposed
. Perrig, “SIA: Secure In-
formation Aggregation in Sensor Networks,” Proceedings
[3] B. Przydatek, D. Song and A
method consumes less energy than the LITEWORP+SEF
method when the FTR is increased because false reports
are detected and dropped in the early routing stages. The
LITEWORP+SEF method consumes more energy than
the proposed method because LITEWORP monitors all
traffic in the network and SEF attaches message authen-
tication codes to all reports. The proposed method uses
ACK messages instead of monitoring traffic. The result-
ing large report size of the LITEWORP+SEF method is
ov
of the First International Conference on Embedded Net-
worked Sensor Systems, Vol. 15, No. 1, 2003, pp. 255-
265. doi:10.1145/958491.958521
[4] A. C. Ferreira, M. A. Vilaca, L. B. Oliveira, E. Habib, H.
C. Wong and A. A. Loureiro, “On the Security of Cluster-
Cased Communication Protocols for Wireless Sensor
Networks,” Lecture Note in Computer Science, Vol. 3420,
2005, pp. 449-458. doi:10.1007/978-3-540-31956-6_53
[5] S. Park, A. Savvides and M. B. Srivastava, “SensorSim:
A Simulation Framework for Sensor Networks,” Pro-
ceedings of International Workshop on Modeling Analy-
sis and Simulation of Wireless and Mobile Systems, 2000,
pp. 104-111.
[6] M. Tubaishat and S. Madria, “Sensor Networks: An
Overview,” IEEE Potentials, Vol. 22, No. 2, 2003, pp.
20-23. doi:10.1109/MP.2003.1197877
[7] F. Ye, H. Luo and S. Lu, “Statistical En-Route Filtering
of Injected
erhead, but the proposed method is unaffected by the
report size.
6. Conclusion
In this paper, we proposed a secure routing method for
detecting false reports and wormhole attacks in wireless
sensor networks. The LITEWORP and SEF methods
have two problems because they are designed without
consideration of each other. The first problem is false
alarms that isolate normal nodes. The SEF method de-
tects false reports and drops them but LITEWORP re-
gards this dropping as malicious behavior and isolates
the node. The second problem is energy consumption;
the report size increases because SEF attaches MACs to
reports, and LITEWORP monitors all of this increased
traffic in th
False Data in Sensor Networks,” IEEE Jour-
nal on Selected Areas in Communications, Vol. 23, No. 4,
2005, pp. 839-850. doi:10.1109/JSAC.2005.843561
[8] H. Yang ans Cipher Based En-
outing Techniques in
S. Lu, “Commutative
Route Filtering in Wireless Sensor Networks,” Proceed-
ings of the 60th IEEE Vehicular Technology Conference,
Vol. 2, 2004, pp. 1223-1227.
[9] J. N. Al-Karaki and A. E. Kamal, “R
Wireless Sensor Networks: A Survey,” IEEE Wireless
Communications, Vol. 11, No. 6, 2004, pp. 6-28.
doi:10.1109/MWC.2004.1368893
thod
thus consumes a lo
solves these proble[10] Z. Yu and Y. Guan, “A Dynamic En-Route Scheme for

H. M. CHOI ET AL. 40
Filtering False Data Injection in Wireless Sensor Net-
works,” Proceedings of the 25th IEEE International Con-
ference on Computer Communications, 2006, pp. 1-12.
[11] S. Zhu, S. Setia, S. Jajodia and P. Ning, “An Interleaved
Hop-by-Hop Authentication Scheme for Filtering of In-
jected False Data in Sensor Networks,” Proceedings of
IEEE Symposium on Security and Privacy, 2004, pp.
259-271.
[12] F. Li and J. Wu, “A Probabilistic Voting-Based Filtering
Scheme in Wireless Sensor Networks,” Proceedings of
the 2006 International Conference on Wireless Commu-
nications and Mobile Computing, 2006, pp. 27-32.
doi:10.1109/ICWMC.2006.5
[13] H. Y. Lee and T. H. Cho, “Fuzzy Adaptive Selection of
Filtering Schemes for Energy Saving in Sensor Net-
works,” IEICE Transactions on Communications, Vol.
E90-B, No. 12, 2007, pp. 3346-3353.
[14] H. Y. Lee and T. H. Cho, “Key Inheritance-Based False
Data Filtering Scheme in Wireless Sensor Networks,”
Lecture Notes in Computer Science, Vol. 4317, 2006, pp.
116-127. doi:10.1007/11951957_11
[15] Y. C. Hu, “Packet Leashes: A Defense against Wormhole
Attacks in Wireless Networks,” IEEE Infocom, 2003, Vol.
3, 2003, pp. 1976-1986.
[16] L. Hu and D. Evans, “Using Directional Antennas to Pre-
vent Wormhole Attacks,” Proceedings of the 11th Annual
Network and Distributed System Security Symposium,
2004, pp. 1-11.
[17] I. Khalil, “LITEWORP: A Lightweight Countermea
for the Wormhole Attack in
sure
in Multi-hop Wire-
Multihop Wireless Net-
works,” Proceedings of International Conference on De-
pendable Systems and Networks, 2005, pp. 612-621.
[18] S. Capkun, L. Buttyán and J.-P. Hubaux, “SECTOR:
Secure Tracking of Node Encounters
less Networks,” Proceedings of the 1st ACM Workshop
on Security of Ad Hoc and Sensor Networks, 2003, pp.
21-32. doi:10.1145/986858.986862