Safety in Wireless Sensor Networks:

Types of Attacks and Solutions

Hctor KASCHEL, Jos MARDONES, Gustavo QUEZADA
Departamento de Ingeniera Elctrica, Facultad de Ingeniera, Universidad de Santiago de Chile,
Avenida Bdo. OHiggins 3363, Estacin Central, Santiago, Chile
hector.kaschel@usach.cl, jose.mardonesf@usach.cl, gustavo.quezada@usach.cl
Abstract: Wireless Sensor Networks (WSN) are highly vulnerable in their security. They are generally deployed in
hostile environments to collect different kinds of data, so they are exposed to serious physical and software attacks.
Therefore, developing methods to increase security becomes an essential aspect of the study of these networks. WSNs are
used in military, medical and biological applications, among others. This paper analyzes the security requirements of
WSNs, the main attacks, and their main characteristics; finally it reviews some security methods currently proposed or
implemented in these networks, and identifies the possible types of security attacks that can affect the layers of the OSI
Reference Model.
Keywords: Wireless Sensor Networks (WSN), Types of attacks.

1. Introduction to satisfy those variable demands for the number

and diversity of products. To fulfil those
Wireless Sensor Networks are formed by a requirements, industry must develop two basic
large number of small nodes, which consume concepts: flexible manufacturing and processes.
low levels of energy and are low cost. These In this context, wireless networks can contribute
sensors are easy to deploy in an area designed process flexibility by offering mobility and
for wireless networks performing specific control independent of the physical location of
functions. The fact that these networks are the process. They also offer control for
formed by a large number of nodes allows them processes in motion or for moving parts; they
to be used in industrial environments to carry offer access to places that are inaccessible to
out the control functions. Their configuration wired networks, and provide safety with respect
makes it possible for the sensors to be located to broken wires. Figure 1 shows the main
abundantly, at all points of interest, and at a applications of wireless sensor networks.
low cost. These kinds of networks also offer the On the other hand, industrial processes must
same services and advantages as normal offer high energy availability and efficiency so
wireless networks. Because of the low cost of as to satisfy the associated production
the nodes, it is possible to reconfigure them or requirements. A lack of availability may mean
adapt them to specific access points, providing great economic loss caused by stopping some
access to a larger network such as, for example, highly important production line, or even worse,
the Internet. the whole production process. Also, the system
Rapid technological advances cause products must have a high level of security so that there is
and services to vary constantly, forcing industry no change in information at any time.

Figure 1. Application of Wireless Sensor Networks

Studies in Informatics and Control, Vol. 22, No. 3, September 2013 http://www.sic.ici.ro 323
Section 2 analyzes the security requirements; how long the message must continue
section 3 shows the different types of security being confidential.
attacks and their impact on the wireless
4. Availability: It encompasses the access to
sensors; section 4 describes some proposed
the information and to the systems by
designs and security solutions used for WSNs;
authorized persons at the time when they
section 5 provides a security approach based on
need it. The introduction of encryption
the layers of the OSI Reference Model; and
algorithms in the wireless sensor networks
finally section 6 gives the conclusions of this
makes them more costly. Whichever
research work.
method is implemented, it increases
processing time, which finally involves
2. Security Requirements greater energy consumption, greatly
threatening the availability of the network.
WSN networks were initially designed for
military purposes, so their field of work was in 5. Data Refreshing: Although the
hostile environments from the beginning. Each confidentiality and the integrity of the data
sensor node is constantly open so they are are ensured, guaranteed message updating
susceptible to attack. The information that is also required. Data refreshing implies
travels through the node has a high probability that the data are up-to-date therefore it is
of being stolen, decoded and used against it. important to ensure that an adversary node
has not replaced the current message by an
The security requirements of a wireless sensor old one.
network [1] can be classified as follows:
1. Authentication: Authentication of 3. Types of Security Attacks on WSNs
messages is fundamental for the different
applications in WSNs, preventing an Table 1 shows the types of attacks, and their
attacker from cloning a node or extracting main characteristics, that can invade a Wireless
password information to send malware to Sensor Network. Attacks on WSNs can be
the network. Therefore, it is necessary to considered from two standpoints: one is the
generate mechanisms that allow the nodes attack to the security mechanisms, and the
to authenticate the information received, other is an attack on the basic mechanisms
which is possible by means of the (routing mechanisms) [2], [3], [4], [5], and [38]
validation of the identity of the as indicated below:
transmitting node.
As seen in Table 1, the attacks can be classified
2. Integrity: This is the property that aims to according to the different modes of action:
keep the data free from unauthorized
Attack techniques:
modifications. Integrity consists in keeping
the accuracy of the information exactly as The attacker can listen to the transmitted
it was generated, without it being packets to analyze the traffic or the
manipulated or altered by unauthorized cryptography.
persons or processes. Data Authentication The attacker can introduce false packets in
can also provide Data Integrity. the network to confuse the sensor network.
3. Confidentiality: Data confidentiality The attacker can introduce malicious nodes
refers to protection of the data as well as to modify the packets received before
of the information exchanged between an resending them.
emitting node and one or more addresses
from third parties. Guaranteeing it requires Node compromise:
a mechanism for communication An attacker can capture a sensor node and
enciphering and occultation. Digitally, the extract all its secrets to use them in the
confidentiality of a document can be security protocols.
maintained through the use of asymmetric
Often, the attacker is not interested in the
keys. The enciphering mechanisms
content of the data, but rather in the
guarantee the confidentiality during the
interference of communications between
time needed to decipher the message. For
the nodes.
that reason it is necessary to determine

324 http://www.sic.ici.ro Studies in Informatics and Control, Vol. 22, No. 3, September 2013
 Table 1. Main Attacks and their Characteristics.

Attacks Characteristics
DoS (Denial of Service) - Produced by the unintentional action of nodes or the action of an
[6] [7],[8] attacker.
Attacks on the in transit - They alter, falsify and repeat the information in transit to the
information [9] [10] source.
- They take control of one node and are capable of manufacturing
new falsified packets.
- Their scope can be over several sensor nodes at the same time.
Sybil attack [11] [13] - An attacker takes the identical characteristics of another node to
become involved in the network
- This attack tries to degrade the integrity of the data, of the security
and the use of the resources that the algorithm of is accessible
through the stolen sensor.
- Attacks the distributed storage, routing mechanisms and data
aggregation.
- When attacked, the network can fight back with strong protocols.
Backhole/Sinkhole - One node acts as a black hole to attract the whole group of sensor
Attack [14] [15] nodes.
- When the malicious node intercepts the communication nodes, it
can do anything with them.
Attack Hello Flood [16] - The attacker uses greetings packets to attract and convince the
[17] nodes.
- The nodes are convinced that the attacker is its neighbor.
- Once the nodes send the packet to the receiver, they must pass
through the attacker, intercepting the packets and doing what they
want it to do.
Wormhole Attack [18], - In this critical attack, the attacker saves the packets in a network
[19] address and the tunnels in another.
- It is a significant threat, because it can occur at the beginning, when
the sensor nodes are just finding out about the neighboring sensors.

An attacker can interfere with the Internal attacks compromise a network

communication channel to deteriorate it, node or they implement malicious nodes.
causing packet loss.
As indicated in Table 1, in the face of the
Passive vs. active: different types of attacks, it is indispensable to
have security mechanisms for WSNs, since
In the passive mode the attacker gets
they are very prone to attacks, and as stated,
information without being discovered.
there are many different kinds of attacks.
In the active mode the attacker is more
aggressive and launches various attacks to 4. Security Methods for the
damage the network.
Wireless Sensor Networks
Internal vs. external:
External attacks are limited and can be In recent years, WSN security has attracted the
launched only from outside the networks attention of a large number of researchers all
application environment. over the world.

Studies in Informatics and Control, Vol. 22, No. 3, September 2013 http://www.sic.ici.ro 325
 Table 2. Summary of different Security Methods applied to Wireless Sensor Networks

Security method Attacks Network architecture Main characteristics

JAM [20], [21] DoS attack Traditional wireless Uses linked neighbouring nodes
sensor networks. to prevent avoidance of the
jammed region.
Based on DoS attack Hybrid sensor network Uses Wormholes to avoid
Wormhole [18], (wireless and wired). jamming.
[19]
Random key pre- Sybil attack. Large number of Uses radio resources, random key
distribution, radio sensors. Highly dense pre-distribution, registration
resource testing, wireless sensor procedure, verification of
etc. [12], [13] network. position, and code testing for
detecting Sybil entity.
Two-directional Hello flood attack. Traditional wireless Adopts a secret, probabilistic,
verification, sensor networks. sharing compartment. Also uses
multi-base station two-directional verification and
routing, multiple-base station routing and
multirouting [16], multirouting.
[22]
Based on Information or Traditional wireless Efficient use of the resources.
communication data spoofing. sensor networks. Protects the network even if part
security [23] of the network is compromised.
TIK[19] Wormhole attack, Traditional wireless Based on symmetric
information or sensor networks. cryptography, requires
data spoofing. synchronization between all
communicating parties,
implements temporary leashes.
Pre-distribution of Data and Traditional wireless Provides resilience in the
random key information sensor networks. network, protects the network,
[24][25] spoofing. Attacks even if part of the network is
information in compromised, provides
transit. authentication measures for
sensor nodes.
Eschenauer & Data and Distributed sensor Allowed for a large number of
Gligor, [26] information network, large scale of wireless sensors that make it
spoofing. wireless sensor possible to add and remove
network with a sensors. Resilient to the capture
dynamic nature. of a sensor node.
REWARD [27] Blackhole attacks. Traditional wireless Uses geographic routing, takes
sensor networks. advantage of being the sender to
see the neighbours transmission
and detects blackhole attacks.
TinySec [28][29] Data and Traditional passive Centred on providing message
information wireless sensor authenticity, integrity and
spoofing, the networks. confidentiality messages-works
messages repeat in the link layer.
the attacks.
SNEP y TESLA Data and Traditional passive Replay protection, semantic
[30][31] information wireless sensor security, data authentication, low
spoofing, the networks. communication overhead.
messages repeat
the attacks.

326 http://www.sic.ici.ro Studies in Informatics and Control, Vol. 22, No. 3, September 2013
Table 2 shows a revision of various proposed radiofrequency transceptor, the signal's
or implemented security methods based on the energy consumption, and the selection of
type of WSN attack [5], and their access channels. Its objective is to increase
main characteristics. reliability, while subsequently reducing
loss. The most frequent types of attacks
The main idea followed by the Wireless Sensor
are DoS.
Networks in the matter of security is to have an
integral approach, so as to improve the Table 3 presents a synthesis of the possible
performance of the networks with respect to attacks that can appear in the different layers of
security, longevity and interconnectivity under the OSI Reference Model of a Wireless Sensor
the changing environmental conditions. Network [1], [32], [33]:
Table 3. Possible attacks on the WSN layers.
5. Security Approach based on
WSN Layer Types of attacks
The Layers of the OSI
Physical Denial of Service (DoS)
Reference Model Data link Jamming
Collision
It is important to carry out a holistic approach
to security in which the OSI Reference Model Network Denial of Service (DoS)
Wormhole
layers participate to guarantee the network's
Sinkhole
overall security. Therefore we have that: Sybil
- In the application layer: The data are Hello flood
collected and administered in the Application Malicious node
application layer; it is important to ensure Desynchronization
the reliability of the data and to transmit
them to the lower levels. One of the 6. Conclusions
problems that can arise is a security
attack on the desynchronization of the Most security attacks in WSNs are caused by the
data transfer. insertion of false data by the compromised
- In the network layer: It is in charge of nodes within the network. This paper presents
identifying the existing routing between the requirements, the different types of security
one or more networks. The objective of attacks, a review of state-of-the-art main security
this layer is to make the data arrive methods proposed or implemented in Wireless
from their origin to their destination, even Sensor Networks, and a security approach based
when both are not directly connected. on the layers of the OSI Reference Model, in
Its aim is to find the best route, making Wireless Sensor Networks.
use of efficient routing algorithms. The
kinds of attacks that can occur in this layer Acknowledgements
are often Wormhole, Sinkhole, Sybil and
Hello Flood. This research was done with the financial
contribution of project DICYT USACH Code
- In the data link layer: This layer is in 061213KC Design and implementation of an
charge of physical addressing of the IWSN (Industrial Wireless Sensor Network)
network's topology, access to the medium, Tolerant to Faults, Energy Efficient, and with
error detection and/or correction, ordered High Security.
mesh distribution, and flow control. This
layer is vulnerable to Jamming and
Collisions attacks that cause collision of
packets and therefore shorter useful life of
the battery by having to retransmit the
packets, generating confusion in the
neighbouring nodes.
- In the physical layer: It is in charge of
providing the data transmission service
over the medium and also controlling the

Studies in Informatics and Control, Vol. 22, No. 3, September 2013 http://www.sic.ici.ro 327
REFERENCES 9. JIA, G., F. JIAN'AN, X. CHEN, Survey on
Secure Data Aggregation for Wireless
1. JAIN, A., K. K. KANT, M. R. Sensor Networks, IEEE International
TRIPATHY, Security Solutions for Conference on Service Operations,
Wireless Sensor Networks, Second Logistics and Informatics (SOLI), 2011,
International Conference on Advanced pp.138-143.
Computing & Communication 10. PFLEEGER, C. P. S. L. PFLEEGER,
Technologies (ACCT), 2012, pp. 430-433. Security in Computing, 3rd edition,
2. MODARES, H., R. SALLEH, A. Prentice Hall 2003.
MORAVEJOSHARIEH, Overview of 11. CHEN, S., G. YANG, S. CHEN, A
Security Issues in Wireless Sensor Security Routing Mechanism Against
Networks, Third International Conference Sybil Attack for Wireless Sensor
on Computational Intelligence, Modelling Networks, International Conference on
and Simulation (CIMSiM), 2011, Communications and Mobile Computing
pp. 308-311. (CMC), Vol. 1, 2010, pp. 142-146.
3. SHARMA K., M. K. GHOSE, D. 12. YI, S., C. YONGFENG, T. LIANGRUI, A
KUMAR, R. PEEYUSH KUMAR, Multi-phase Key Pre-distribution
SINGH, V. KUMAR PANDEY, A Scheme based on Hash Chain, 9th
Comparative Study of Various Security International Conference on Fuzzy Systems
Approaches Used in Wireless Sensor and Knowledge Discovery (FSKD), 2012,
Networks. In IJAST, vol. 7, April 2010. pp. 2061-2064.
4. JAYDP, S., A Survey on Wireless Sensor 13. NEWSOME, J., E. SHI, D. SONG, A.
Network Security, Technical Report 55- PERRIG, The Sybil Attack in Sensor
77, International Journal of Networks: Analysis & Defences,
Communication Networks and Proceedings of the Third International
Information Security (IJCNIS) Vol. 1, N2 Symposium on Information Processing in
August 2009. Sensor Networks, ACM, 2004,
5. PATHAN, A. S. K. H.-W. LEE, C. S. pp. 259-268.
HONG, Security in Wireless Sensor 14. SHARMILA, S., G. UMAMAHESWARI,
Networks: Issues and Challenges, The Detection of Sinkhole Attack in Wireless
8th International Conference on Advanced Sensor Networks Using Message Digest
Communication Technology, ICACT 2006, Algorithms, International Conference on
vol. 2, 2006, pp. 1043-1048. Process Automation, Control and
6. DAOJING, H., C. CHEN; CHAN, S. Computing (PACC), 2011, pp. 1-6.
JIAJUN BU, DICODE, DoS-Resistant 15. CULPEPPER, B. J. H. C. TSENG,
and Distributed Code Dissemination in Sinkhole Intrusion Indicators in DSR
Wireless Sensor Networks, IEEE MANETs, Proceedings of the First
Transactions on Wireless Communications, International Conference on Broadband
vol. 11, Issue 5, 2012, pp. 1946-1956. Networks, 2004, pp. 681-688.
7. OUYANG, X., T. BIN, L. QI, Z. JIAN-YI, 16. WANG, W., J. XU, J. WANG, Detection
H. ZHENG-MING, X. YANG, A Novel and Location of Malicious Nodes based
Framework of Defence System Against on Source Coding and Multi-path
DoS Attacks in Wireless Sensor Transmission in WSN, 11th IEEE
Networks, 7th International Conference on International Conference on High
Wireless Communications, Networking and Performance Computing and
Mobile Computing (WiCOM), 2011, Communications, 2009, pp. 458-463.
pp. 1-5.
17. KARLOF, C., D. WAGNER, Secure
8. WANG, B-T. H. SCHULZRINNE, An IP Routing in Wireless Sensor Networks:
traceback mechanism for reflective DoS Attacks and Countermeasures, Elsevier's
attacks, Canadian Conference on Electrical Ad Hoc Network Journal, Special Issue on
and Computer Engineering, Volume 2, 2-5 Sensor Network Applications and
May 2004, pp. 901-904. Protocols, September 2003, pp. 293-315.

328 http://www.sic.ici.ro Studies in Informatics and Control, Vol. 22, No. 3, September 2013
18. HARBIN, J., P. MITCHELL, D. PEARCE, 26. ESCHENAUER, L., V. D. GLIGOR, A
Wireless Sensor Network Wormhole Key-management Scheme for
Avoidance using Reputation-based Distributed Sensor Networks,
Routing, 7th International Symposium on Proceedings of ACM CCS'02, 18-22
Wireless Communication Systems November 2002, pp. 41-47.
(ISWCS), 2010, pp. 521-525.
27. KARAKEHAYOV, Z., Using REWARD
19. TRIKI, B., S. REKHIS, N. BOUDRIGA, to Detect Team Black-hole Attacks in
Digital Investigation of Wormhole Wireless Sensor Networks, in Workshop
Attacks in Wireless Sensor Networks. on Real-World Wireless Sensor Networks
Eighth IEEE International Symposium on (REALWSN'05), 20-21 June, 2005,
Network Computing and Applications, Stockholm, Sweden.
2009, pp. 179-186.
28. MOON, M., D. S. KIM, JONG SOU
20. XUAN, Y., Y. SHEN. N. P. NGUYEN, M. PARK, Toward Modelling Sensor Node
T. THAI, A Trigger Identification Security Using Task-Role Based Access
Service for Defending Reactive Jammers Control with TinySec, Conference on
in WSN, IEEE Transactions on Mobile Computational Intelligence and Security,
Computing, vol. 11, Issue 5, 2012, vol. 2, 2006, pp. 1109-1112.
pp. 793-806.
29. KARLOF, C., N. SASTRY, D. WAGNER,
21. MAHMOOD, A. R., H. H. ALY, M. N. TinySec: A Link Layer Security
EL-DERINI, Defending Against Energy Architecture for Wireless Sensor
Efficient Link Layer Jamming Denial of Networks, 2 International Conference on
Service Attack in Wireless Sensor Embedded Networked Sensor Systems,
Networks, 9th IEEE/ACS International Baltimore, MD, USA, 2004, pp. 162-175.
Conference on Computer Systems and
30. LI, Y., F. LIU, L. DING, Research about
Applications (AICCSA), 2011, pp. 38-45.
Security Mechanism in Wireless Sensor
22. HAMID, M. A., M.-O. RASHID, C. S. Network, International Conference on
HONG, Routing Security in Sensor Image Analysis and Signal Processing
Network: Hello Flood Attack and (IASP), 2011, pp. 447-451.
Defence, to appear in IEEE ICNEWS
31. YEO, D.-G., H.-Y. YOUM, An TESLA
2006, 2-4 January, Dhaka.
Protocols with Multi-senders Based on a
23. SLIJEPCEVIC, S., M. POTKONJAK, V. 2-Level XOR Chain with Data-Loss, 10th
TSIATSIS, S. ZIMBECK, M. B. International Symposium on Tolerance
SRIVASTAVA, On Communication Applications and the Internet (SAINT),
Security in Wireless Ad-hoc Sensor 2010, pp. 269-272.
Networks, 11th IEEE International
32. PANDEY, A., R. C. TRIPATHI, A Survey
Workshops on Enabling Technologies:
on Wireless Sensor Networks Security,
Infrastructure for Collaborative Enterprises,
International Journal of Computer
10-12 June 2002, pp.139-144.
Applications, Vol. 3, N 2, June 2010, pp.
24. DU, W., J. DENG, Y. S. HAN, P. K. 43-49.
VARSHNEY, A Pairwise Key Pre-
33. KASCHEL, H., Y. B. L. SANCHEZ, J.
distribution Scheme for Wireless Sensor
MARDONES, G. QUEZADA, Modelling
Networks, Proceedings of the 10th ACM
Sensor Node Security Using Task-Role
Conference on Computer and
based Access Control with TinySec,
Communications Security, 2003, pp. 42-51.
Studies in Informatics and Control Journal.
25. CHAN, H, A. PERRIG, D. SONG, Vol. 20, N 3, September 2011,
Random Key Predistribution Schemes pp. 285-292.
for Sensor Networks, In IEEE Symposium
on Security and Privacy, Berkeley,
California, May 11-14 2003, pp. 197-213.

Studies in Informatics and Control, Vol. 22, No. 3, September 2013 http://www.sic.ici.ro 329