30-08-2014, 02:40 PM
Review on a packet hiding: a new paradigm for avoiding jamming attack over wireless network Project Report
Review on a packet.pdf (Size: 427.76 KB / Downloads: 21)
ABSTRACT
The wireless network are usually preferred because of its challenging features such as its faster accessibility,
compatibility and its connectivity among extended variant set of users. Due to its better transfer rate the
authentication mechanism is ignored in wireless sensor network. This shortens the limitation of the existing
wired network. By using the wireless sensor network various types of jamming attacks are invited. Some
detection strategies are available but they are failed sometime in analyzing and reporting the presence of
jammer. In case of external threat model it is quite easy but in case of an internal threat model the person has
great knowledge about network secrets and internal protocol specifications, so these persons can easily launch
low effort squeeze attacks. These attacks are very difficult to detect and count. So as to protect the packets from
such attacks various methodologies are implemented. The main goal of these methodologies is to prevent that
packet from the jammer, so as to enable the safe transmission among intended nodes even in presence of
jammer
INTRODUCTION
In wireless network due to its openness nature various intentional interference attacks are comes into existence.
Anyone with a transceiver can easily launch jamming attack in an ongoing transmission, or create noise or collision or block
the transmission of authorized one or inject spurious messages. One of the best way to degrade the network performance is
to jam the wireless transmission or to allow the adversary to take more efforts on jamming the network [1,2]. In the simplest
form of jamming, the adversary takes first few bytes of transmitted packet classifies it and corrupt it by causing
electromagnetic interference such as magnetic radio waves, FM modulated noise in the network‟s operational frequencies,
and in proximity to the targeted receivers. Under this strategy, jammer includes either continuous or random transmission of
high interference signals [3], but due to this it has several disadvantages occurred. The first is that the adversary has to spend
its more amount of energy to jam frequency bands of interest and the second one, Due to continuous presence of unusually
high interference levels make these types of attacks easy to detect [1,2], [4].
The adversary considered in proposed system is active but only for short period of time. These adversaries target
the messages which have more importance. The examples are rout request messages, rout reply messages or the TCP
acknowledgement [5]. So the first step of the attacker is that he must be capable of implementing the strategy called as
“Classify then jam” before wireless transmission completes.
Suppose there are two communicating parties A (sender) and B (receiver) and J is the jamming node within their
communication range. Now A sends packet m to B, the goal of J is takes first few bytes of m classifies them and then corrupt
these few bytes. And then visualize to A as J is nothing but B and then start to communicate with A. In this way J requests
for more and more packets to A in order to target them for attacking purpose. In this way jammer attacks but the main
condition required for the attacker is that he must be knowledgeable at every layer of the TCP protocol. The whole
communication is shown in figure 1which gives the actual study of jamming attack
RELATED WORK
In related work we are going to study the reasons for jamming, the requirements for it, whether it
happens intentionally or intentionally. At the last we are going to study about spread spectrum technique which
is used by conventional anti-jamming methods and the disadvantages of existing system, the mechanism for
prevention they used.
Due to jamming wireless transmission either stopped or disturbed. This jamming is either in the form
of interference, noise or collision. If the jamming is intentionally then it is in the form of attack otherwise it is
caused due to network load. No any special hardware required for executing it. Conventional anti-jamming
techniques are based on either some form of jamming evasion or SS communications [3], [9]. Now we are going
to study in detail about Spread Spectrum (SS) communications.
Spread spectrum communications works as follows. First the input is given to channel encoder, now
the channel encoder creates analog signal which consist of narrow bandwidth. Now this generated signal is
modulated by using the sequence of digits. Pseudo noise or pseudo-random number generator is the main source
for generating the Spreading code. The main reason for using „modulation‟ mechanism is to increase signal‟s
bandwidth which is going to be transmitted. This whole procedure is carried at the sender side. Now at the
receiver side for demodulating the spread spectrum signal digital sequence is used. This generated signal is
given to channel decoder in order to recover the original data. Spread spectrum is generally used for hiding and
encrypting signals.
PROBLEM STATEMENT AND ASSUMPTIONS
Under this chapter we are going to study the actual problem statement of the proposed system and the
various kinds of assumptions made by the system for the understanding purpose.
Problem Statement
Consider the scenario depicted in Figure 1. Nodes A and B communicate via a wireless link. Within the
communication range of both A and B, there is a jamming node J. When A transmits a packet m to B, node J
classifies m by receiving only the first few bytes of m. J then corrupts these few bytes by interfering between
their communication and viewing A that J is nothing but B and then start to communicate with A for corrupting
the messages received from A. We address the problem of preventing the jamming node from classifying m in
real time, thus mitigating J‟s ability to perform selective jamming. The main goal of the proposed system is to
transform a selective jammer to a random one.
Network Model
The network consists of collection of nodes connected via wireless links. In our proposed system we are
considered 12 nodes that are communicate directly if they are within their communication range or they can use
multiple hops. The communication of these nodes is carried out by various types of modes. They can
communicate in unicast mode by using simplex link or bicast mode by using duplex link or in broadcast mode.
Communication is kept either in encrypted format or in unencrypted format depending on the prevention scheme
used. In case of broadcast communication the symmetric keys are shared among all intended receivers which are
generated by asymmetric cryptography or pre-shared pairwise keys.
IMPACT OF SELECTIVE JAMMING
In this section we are going to illustrate the impact of selective jamming on the network performance.
To implement jamming attacks, the proposed system uses Network Simulator 2.34. Here the two wireless
network scenarios are considered. In the first scenario the attacker targeted TCP connection especially at the
transport layer. The second scenario, the jammer focuses on network-layer control messages which are
transmitted during the route establishment process.
Selective Jamming at the Transport Layer
In this experiment, we set up transfer of 3 MB file between two users A and B connected via multi-hop route. The
TCP protocol was used for reliable connection. The RTS/CTS mechanism was enabled at the MAC layer. Here the
transmission rate 11Mbps was considered at each link. The jammer was placed within communication range of A and B.
various selective jamming strategies were considered here, 1.Selective jamming of cumulative TCP-ACKS. 2. Jamming of
any data packet. 3. Random jamming. 4. Jamming at MAC layer means RTS/CTS messages. In each of these strategies
fraction p is jammed. This fraction p is of the targeted packet.
Selective Jamming at the Network Layer
In this scenario, 12 nodes of multi-hop wireless networks are simulated randomly within the square
area. Here AODV routing protocol is used in proposed system for establishing and discovering the routing path
for the data packets. Connections are established between source/destination pairs. In this scenario the jammers
are placed in non-overlapping area of the network. In proposed system continuous, random and targeted RREQ
these types of jammers are considered that block the fraction p of targeted packet. Constant jamming attack is
equally effective to attack on RREQ messages. However, selective jamming is several orders of magnitude more
efficient. But due to flooding process of AODV random jammer fails to disturb route paths
PROPOSED SYSTEM
This chapter describes the brief introduction of proposed system. The way that proposed system
provides security and the advantages of it over existing systems. Proposed system provides intuitive solution
against jamming by encrypting the entire packet along with the header. For generating the cipher text static key
is used. The static decryption key is shared among all intended receivers in case of broadcast communication.
Though we kept encryption key secret, but static portion of packet which has been transmitted can be used for
packet classification. This is one of the advantages of proposed system.
The main advantages of proposed systems are- It is very easy for exploiting knowledge of protocol and
cryptographic primitives extracted from compromised nodes. The other main advantage is that the proposed
system shows that selective jamming attacks lead to DOS by taking very less effort on behalf of jammer. In this
way the proposed system achieves strong security protocols.
METHODS USED
There are various methods that are used for achieving strong security of system we will go through
them in detail. We will first study how the jammer performs attack on the system with the help of real time
packet classification then we will further study the various cryptographic schemes that prevent these jamming
attacks.
IMPLEMENTATION STEPS
The figure 5 shows detail implementation of the proposed system with its analysis and throughput.
There are following steps carried out to perform the implementations these are as follows.
Implementation of wireless node in NS-2 with AODV.
Implementation of jamming attack with selective transmission.
Implementation of packet classification for wireless traffic.
Implementation of packet hiding for real packet.
Detection of jamming attack and analysis with throughput
CONCLUSION
This paper solves the problem of selective jamming attacks in wireless network. Here an internal threat
model is considered in which the jammer is part of the network, thus being aware of the protocol specifications
and network secrets. Jammer can classify transmitted packet in real time by decoding the first few symbols of an
ongoing transmission or packet. We evaluated the impact of selective jamming attacks on network protocol such
as on TCP and routing. This paper developed three schemes that transform a selective jammer to a random one
by preventing real time packet classification. These schemes combine cryptographic primitives such as
commitment schemes, cryptographic puzzles, and all-or-nothing transformations with physical-layer
characteristics. We analyze the security of our method and quantified their computational and communication
overhead