Denial of Service (DoS) Attacks at Network Layer in WSN

Sunil Ghildiyal et al. Int. Journal of Engineering Research and Applications www.ijera.com
ISSN: 2248-9622, Vol. 5, Issue 11, (Part - 1) November 2015, pp.86-89
www.ijera.com 86 | P a g e
Denial of Service (DoS) Attacks at Network Layer in WSN
Sunil Ghildiyal, Bhupender Singh Rautela, Anupam Semwal
Uttaranchal University Dehradun Uttarakhand
Graphic Era University Dehradun Uttarakhand
Drona College of Mgmt & Tech. Ed. Dehradun Uttarakhand
ABSTRACT
Recent advancements in technology, tiny size, cost effectiveness have made sensors as a crucial part of real
world sensitive applications. These sensor nodes are scattered over an area to monitor the situations like fire,
flood and record the data and to forward meaningful data to the center head node for observation, resulting an
advance prompt to control the situation. In last decade, WSN have grown significantly in variety of areas and
applications, resulted the high, consistent security mechanism. Also, there is variety of attacks on WSN at their
different layers of architecture. Though sensor nodes are not capable enough in terms of power, processing etc.
but applications based on these sensors demand on-time collection of information or data and then to send same
on reliable, secure delivery medium. Small sensors with limited hardware, processing cannot afford traditional
security mechanisms to face or sustain the attacks. There is variety of attacks at different layers of WSN
architecture to affect sensor‟s roles like signaling, framing, transmission etc. Many Denial of Service (DoS)
attacks are identified at each layer of WSN which are purposeful, planned attacks to hamper the availability of
service, restricting the sensor node‟s utility for problem solution. In this paper we will focus on the WSN
architecture, characteristics, constraints and various types of DoS attacks primarily on physical and data link
layer and particularly at network layer in details with some suggestions against attacks.
KEYWORDS: Wireless, Sensor, Power, Denial of Service DoS, Attack, Vulnerabilities.
I. INTRODUCTION
As a result of recent developments in wireless
technology polishing, wireless networks are now
believed as a reliable architecture medium to deliver
communication with major security parameters
confidentiality, integrity and availability and non-
repudiation. Wireless Sensor Networks consist of less
power, less processing capability, small size sensor
nodes[1]. Hence, It becomes very tough to raise the
capability level of such tiny sensors due to their
various constraints. Constraints, associated with
sensors are to be considered seriously while
designing a secure real world problem solution using
WSN.
Actually, sensor nodes use RF for messaging,
communication and hence use broadcast basically.
Since medium is open, it is tedious to protect the
broadcast from easy eavesdropping, as injecting can
be done very easily over wireless broadcasting. Also,
sensor nodes are scattered over a geographical area in
physically insecure pattern can be stolen easily, can
be tempered physically or replayed or reprogrammed
after capturing. Insecure, open deployment of sensor
nodes make them to be easily detected for damage
purpose[2]. These limited power capacity nodes
make WSNs. very weak and paralyzed architecture in
front of any intended attack like flooding or replaying
etc. One of initial measure against these threats may
be authorization access checklist available with them
to detect unauthorized or malicious users.
II. WSN CHARACTERISTICS
For last about two decades, WSNs. have received a
lot of interest by the researchers, industry. This is
cause of those to be less cost solutions to many real
world problem solving applications. Other favoring
factors are easy to use, low energy consuming nodes,
portability, unattended operation even in no men land
with an ability to withstand bad geographical,
environmental situations, having dynamic network
topology as per situation, faster recovery methods or
alternates with sensor node stopping and failures,
Mobility of nodes, Heterogeneity of nodes and at all
highly scalable in terms of topology and deployment.
III. WSN CONSTRAINTS
Resource: Sensors are equipped with less
capable processors and very low RF linking
bandwidth. Of course, It is due to tiny size and low
battery. Hence, computational capabilities are also
affected by battery and processors.
Memory: A sensor node consists of a flash
memory and flash RAM. But loading of operating
system and other system applications consume much
space, leading less space for other tasks and storage.
In sensors, flash memory is used for storing
downloaded application code.
Message Size: As compared to any traditional
network, message size of WSN is quite small which
results in no concept of segmentation in WSN
applications usually.
RESEARCH ARTICLE OPEN ACCESS

Absence of Global Addressing: As the number
of sensors is very large, hundreds or thousands in an
application, it is not possible to identify each node
with unique addressing at global level.
Location Management: As nodes are small and
scattered in an open area, dislocation of nodes by
environmental conditions like earthquake or
avalanche, mobility of nodes may result in locating
the nodes. It affects data to be collected by the nodes
after they have been deployed at specific place or
have been constant static at same place for a time
period.
Data Redundancy: There is very high chance of
data redundancy as many nodes may capture the
same data of same phenomenon.
Data Availability: It means whether sensor node
has the capability to use the resources of network and
whether the network is available with the messages to
communicate. In WSN, failure of base station or
cluster head‟s availability will also lead to threaten
the entire sensor network. Hence it is important to
maintain a proper operational network.
Self-Organization: A wireless sensor network is
a typically an ad hoc network, where every sensor
node is to be independent and flexible enough for
self-organization and self-healing under different
conditions. WSNs. form random infrastructure as per
situation and need network management in a sensor
network by nodes themselves.
IV. WSN SECURITY REQUIREMENTS
Ultimate goal of security architecture is to
prevent the information from various attacks.
Security measures make sure that the services would
be available even in presence of DoS attacks or any
other vulnerability. It will make sure that only
authorized node can be a part of communication of
information. As result of it, a malicious node cannot
masquerade as trusted node. When authorized users
or nodes are exchanging the information,
confidentiality must be maintained with data
integrity. Data freshness and non-repudiation is also
to be considered as main parameters of security
measures, whether already applied or to be. WSN
nodes are very small and light weight and generally
deployed randomly, operated in unattended
environment subsequently. Hence the security
requirements include self-organization of node which
further defines self-configuration, self-management
(autonomous) and self-healing (fault tolerant).
V. THREAT MODEL
In WSN, threats are from outside the network
and within the network. If attacks are from the nodes
of the native network then it is much harmful. Also, it
is quite difficult to find out the malicious or
compromising node within the native network.
Another classification of the attacks may be passive
and active where passive attacks don‟t modify or
alter the data as active attacks do. If the opponent
attack by using similar capacity nodes for network
penetration it is called mote class attack but when
powerful devices like laptop are used to penetrate the
network then such attack is called laptop attack.
VI. WSN ATTACKS
Attacks on WSN can be divided into two
categories: invasive and non-invasive. Non-invasive
attacks generally target to timings, power and
frequency of channel, trying to destroy the signaling
system. Whereas invasive type attacks aim to hamper
the availability of service, information transition,
routing etc. DoS attack aims the system to be
inaccessible. However during the transit of
information, more common attacks are encountered.
Attacks affect the routing schemes, routing tables and
routing algorithms mostly in general.
VII. DOS AND DOS ATTACKS
There are varieties of DoS conditions, which
may temper the nodes and network operations
subsequently. These attacks may hinder the routines
of the network, may lead to the resource exhaustion,
any software bug, or any difficulty while working
with any application or infrastructure. Such obstacles
in network functionality are called Denial of Service
(DoS) due to its direct affect on availability or fully
functionality of service. But when these are because
of planned intention of the opponent, these are called
DoS attacks.
Dos attacks are intended attack of opponent to
destroy the entire network components or operations.
DoS attack may limit the network operations more
than expected. DoS attack may occur at every layer
of OSI layers of WSN [3]. DoS attacks penetrate the
efficiency of aimed networks by affecting its
associated protocols. DoS attacks may consume or
exhaust the resources, alter the infrastructure
configuration and can demolish the network
components either partial or full.
Wood and Stankovic presented layer wise
categorization of DoS attacks first. [4], which was
further enhanced by Raymond and Midkiff with some
addendums [5]. In this paper, we will discuss about
the DoS attacks at different layers of WSN
infrastructure in general and then we will conclude
with DoS attacks at network layer. .
VIII. DOS ATTACK AT PHYSICAL
LAYER
Jamming is one of attack at physical layer, in
which radio frequencies used by the network nodes
are interfered, adversary can either disrupt entire
network which depends on the power of jamming
nodes. Jamming is of various types Constant,

Deceptive, Random and Reactive [6]. Jamming may
be consistent or intermittent.
Another attack at physical layer is tempering, in
which attacker may physically temper the nodes and
can compromise with them. Temper-proof physical
packaging is one alternate of this attack but costs a
lot[7].
IX. DOS ATTACK AT LINK LAYER
Exhaustion (Continuous Channel Access is one
of major attack where attacker may disrupt the
channel by frequently requesting and transmission
over it. It results in starvation for channel access for
other nodes.
Collision occurs when two nodes intend for same
frequency channel transmission simultaneously.
Attackers may need to induce a collision instance in
one octet of transmission to disturb entire packet
transmission.
Unfairness is also one of attack at data link layer
which is referred as repeated collision based attack or
an abusive use of cooperative MAC layer priority
mechanisms.
X. DOS ATTACK AT NETWORK
LAYER
X.A. False Routing or Spoofed, Replayed Routing
Information
Main focus of such kind of attack is on routing
protocols, specifically on routing. Sensor nodes
exchange routing information at pre determined time
intervals or as per algorithm policy of routing. A
malicious node can change routing information,
resulting to alter the routing of entire WSN
infrastructure or its any partition. This is possible
through altering or changing the routing information,
by narrowing or extending the routing information in
the table or by fake error messages generation. One
of best strategy against such attack is to implement
MAC code with the message. In addition to it, time
stamps can also be added to prevent against replaying
the routing.
X.B. Selective Forwarding
Fundamental principle of sensor network is
„Multi-hop”. It is to ensure that each sensor nodes
will forward the entire message to next node in line
what they received. In selective forwarding, nodes
selectively drop few messages instead of forwarding
everything. Attacking nodes deny routing few
certain messages and drop them. This attack is
effective specially if combined with an attack trying
to collect most of the traffic via node. Sensor
networks assume that nodes faithfully forward
messages what they have received. But some
compromised node might refuse to forward packets
selectively. As result of it neighbor nodes may opt for
alternate route [8].
If all the packets are denied for forwarding by
anode after receiving, is called black hole attack. In
selective forwarding few messages are dropped and
few are forwarded further to the next node. One of
the defense mechanism against this attack is multiple
paths to send the data.
X.C. Sinkhole Attacks
In this attack attackers seem to be more attractive
to its nearby nodes by forging the routing
information. Main aim of such attack is to tempt all
the neighbor nodes. A sinkhole attack tries to lure
mostly all the network traffic toward the
compromised node, creating a metaphorical sinkhole
with the adversary at the center. Geo-routing
protocols are one of the routing protocol classes
which are resistant to sinkhole attack, because that
topology is based on localized information only and
all traffic is naturally routed through the physical
location of the sink node [9].
X.D. Sybil Attack
In this attacker attacks a single node in the
network with a malevolent code masked with
multiple identities or a node duplicates itself and
presented in the multiple locations. Then this node
acts a polymorphic behavior, misleading to others
with multiple identities. Such identities may decrease
topology maintenance, disparity in storage and
routing or targeting fault tolerant systems. This
attack includes a major concern for Geographical
Routing Algorithms which needs the location of a
node to route the message efficiently. Various
authentication and encryption mechanism can prevent
an outsider to launch a Sybil attack on the sensor
network [10].
X.E. Wormhole
Wormhole is referred as low latency link
between two portions of a WSN network over which
an attacker replays network messages [11]. An
adversary can tunnel messages received in one
network partition over a low latency link and replay
them in another partition. In such attack, an
adversary convinces the nodes which are multi hop
away that they are closer to the base station.. The
wormhole attack generally involve two far away
malevolent codes conspire to minimize their
remoteness by replaying packets next to an out-of-
reach channel, is only available to attacker.
X.F. Hello Flood
Many protocols require hello packets to
announce their neighbors about their state and
presence or absence. This attack exploits Hello
packets. Malicious nodes sometime can cause of
immense traffic of useless messages. It is known as
flooding. Malicious nodes, sometime replay some

broadcast traffic which is useless but congest the
channel. In hello flood type attack, attackers use very
high power RF transmitters to handle the large area
of nodes into trusting that they are neighbors of it.
Attacker may also broadcast a fake superior route so
that other nodes will attempt very far from it in RF
distance. Authentication is the solution to such
attacks. Such attacks can easily be avoided by verify
bi-directionality of a link.
X.G. Acknowledgment Spoofing
Many routing algorithms used for WSNs require
transmission of acknowledgment packets from
receiver to sender as a token of successful receipt.
Attacking node may spoof the acknowledgements of
overheard packet destined for neighboring nodes in
order to provide false information to those nodes.
X.H. Node Capture
It is experienced that only single node capture is
also more than sufficient for an attacker to take over
the entire network and doing malicious action to
destroy the network operations.
XI. CONCLUSION
There are various attacks to hamper the smooth
functioning of wireless sensor networks like denial of
sleep, homing etc. In many situations, attacks may
overlap also with each other. It is difficult to measure
the attacks and their solution at physical layer as
sensors have native radios of very low power and is
operated in open area, unattended environment,
hence are very poor to resist such attacks. Though
there are algorithms and security mechanisms for
network security and protection from above attacks
but can not be applied in WSN nodes due to node‟s
constraints. There is need of tiny low computational
algorithms for WSN. However there are many
algorithms existing for WSN infrastructure and being
applied also. But those are failure to be proved as
correct and fruitful measures against above attacks.
DoS situation at any layer in WSN requires to be
addressed by strong mechanism. It is recommended
to develop a prevention scheme against attacks which
can be applied already to make WSNs. much stronger
against DoS attacks. DoS may appear as individual
and sometime altogether. It is always advisable to
develop and deploy a proper suitable measure in
WSN as prevention already.
REFERENCES
[1.] D. K. Chaitanya “Analysis of DoS attacks
on WSN using simulation” Middlesex
University.
[2.] Ritu Sharma et. al. “Analysis of security
protocols in wireless sensor networks” Int.
Journal Advanced Networking and
Applications Vol 02, Issue 03.
[3.] Al-sakib Khan Pathan “Denial of Service in
Wireless sensor networks: issues and
challenges.”Advances in communications
and Media Research ISBN 978-1-60876-
576-8.
[4.] 4. Wood, A. D. and Stankovic, J.A. (2002) “
Denial of Service in Sensor Networks”
IEEE Computer, vol. 35, no. 10, 2002, pp
54–62.
[5.] Raymond, D. R. and Midkiff, S. F. (2008)
“Denial-of-Service in Wireless Sensor
Networks: Attacks and Defenses” IEEE
Pervasive Computing, January-March 2008,
pp 74-81.
[6.] Xu, W., Trappe, W., Zhang, Y., and Wood,
T. “The Feasibility of Launching and
Detecting Jamming Attacks in Wireless
Networks‟ ACM MobiHoc‟05, May 25–27,
2005, Urbana-Champaign, Illinois, USA, pp
46-57.
[7.] Anthony D. Wood, John A. Stankovic
“Denial of Service in Sensor networks”
University of Virginia 0018-9162/02/$17.00
©2002 IEEE
[8.] C. Karl of and D. Wagner, “Secure routing
in wireless sensor networks: Attacks and
countermeasures”, In Proceedings of the 1st
IEEE International Workshop on Sensor
Network Protocols and Applications, May
2003, pp. 113-127.
[9.] Chaudhari H.C. and Kadam L.U.” Wireless
Sensor Networks: Security, Attacks and
Challenges” International Journal of
Networking Volume 1, Issue 1, 2011, pp-
04-16
[10.] Dr. G. Padmavathi, et.a l “A Survey of
Attacks, Security Mechanisms and
Challenges in Wireless Sensor Networks”
(IJCSIS) International Journal of Computer
Science and Information Security,Vol. 4, No.
1 & 2, 2009
[11.] C. Karlof and D. Wagner, “Secure routing in
wireless sensor networks: Attacks and
countermeasures”, In Proceedings of the 1st
IEEE International Workshop May 2003,
pp. 113-127.
[12.] Wang, Q., Zhu, Y. “Reprogramming
Wireless Sensor Networks: Challenges and
Approaches” IEEE Network, May/June
2006.

Denial of Service (DoS) Attacks at Network Layer in WSN

More Related Content

What's hot (16)

Viewers also liked (20)

Similar to Denial of Service (DoS) Attacks at Network Layer in WSN (20)

Recently uploaded (20)

Denial of Service (DoS) Attacks at Network Layer in WSN