Secure Self Re-Organizing of Nodes Using Closeness Technique in Cluster MANET
©2017
Textbook
186 Pages
Summary
Mobile ad hoc network (MANET) is defined as a self-configuring infrastructureless network used for communication by wireless links with the support of mobile devices. A MANET is referred to as a wireless network with independent nodes moving freely with respect to each other. Due to the independent free moves of nodes, a huge amount of packet data loss occurs in transmitting the packet from source to destination. The risk of node misbehaviour is extremely high. The unsecured ad hoc network environment is initiated due to the active nature of networks and node mobility. In addition, the task of key management is more complex in ad hoc network. Due to the nature of free moving characteristics, MANET faces improper node cooperation. The main reason behind ineffective node cooperation is presence of malicious or selfish nodes. Moreover, the existence of malicious unauthenticated nodes causes insecure communication. Hence, the proposed system aims in the development of proper node cooperation, malicious node detection and secure communication in MANET.
Excerpt
Table Of Contents
Sathiya Kumar, C., Duraiswamy, K.: Secure Self Re-Organizing of Nodes Using
Closeness Technique in Cluster MANET, Hamburg, Anchor Academic Publishing 2017
Buch-ISBN: 978-3-96067-185-5
PDF-eBook-ISBN: 978-3-96067-685-0
Druck/Herstellung: Anchor Academic Publishing, Hamburg, 2017
Bibliografische Information der Deutschen Nationalbibliothek:
Die Deutsche Nationalbibliothek verzeichnet diese Publikation in der Deutschen
Nationalbibliografie; detaillierte bibliografische Daten sind im Internet über
http://dnb.d-nb.de abrufbar.
Bibliographical Information of the German National Library:
The German National Library lists this publication in the German National Bibliography.
Detailed bibliographic data can be found at: http://dnb.d-nb.de
All rights reserved. This publication may not be reproduced, stored in a retrieval system
or transmitted, in any form or by any means, electronic, mechanical, photocopying,
recording or otherwise, without the prior permission of the publishers.
Das Werk einschließlich aller seiner Teile ist urheberrechtlich geschützt. Jede Verwertung
außerhalb der Grenzen des Urheberrechtsgesetzes ist ohne Zustimmung des Verlages
unzulässig und strafbar. Dies gilt insbesondere für Vervielfältigungen, Übersetzungen,
Mikroverfilmungen und die Einspeicherung und Bearbeitung in elektronischen Systemen.
Die Wiedergabe von Gebrauchsnamen, Handelsnamen, Warenbezeichnungen usw. in
diesem Werk berechtigt auch ohne besondere Kennzeichnung nicht zu der Annahme,
dass solche Namen im Sinne der Warenzeichen- und Markenschutz-Gesetzgebung als frei
zu betrachten wären und daher von jedermann benutzt werden dürften.
Die Informationen in diesem Werk wurden mit Sorgfalt erarbeitet. Dennoch können
Fehler nicht vollständig ausgeschlossen werden und die Diplomica Verlag GmbH, die
Autoren oder Übersetzer übernehmen keine juristische Verantwortung oder irgendeine
Haftung für evtl. verbliebene fehlerhafte Angaben und deren Folgen.
Alle Rechte vorbehalten
© Anchor Academic Publishing, Imprint der Diplomica Verlag GmbH
Hermannstal 119k, 22119 Hamburg
http://www.diplomica-verlag.de, Hamburg 2017
Printed in Germany
i
ABSTRACT
Mobile ad hoc network (MANET) is defined as a self-configuring
infrastructureless network used for communication by wireless links with the
support of mobile devices. A MANET is referred as wireless network with
independent nodes moving freely with respect to each other. Due to the
independent free moves of nodes, a huge amount of packet data loss occurs in
transmitting the packet from source to destination. The risk of node
misbehavior is extremely high. The unsecured ad hoc network environment is
initiated due to the active nature of networks and node mobility. In addition,
the task of key management is more complex in ad hoc network. Due to the
nature of free moving characteristics, MANET faces improper node
cooperation. The main reason behind ineffective node cooperation is presence
of malicious or selfish nodes. Moreover, the existence of malicious
unauthenticated nodes causes insecure communication. Hence, the proposed
system aims in the development of proper node cooperation, malicious node
detection and secure communication in MANET.
In order to enhance the node cooperation, and malicious node
detection, first research work develops a Secure Key Model (SKM). The main
idea behind SKM is to cluster the nodes based on reputation and ranking.
Clustering of node is reorganized on its own self with the evaluation co-
operative nodes. The negative reputation value in reputation table detects
malicious node. Performance evaluations are carried out. SKM achieves the
high performance rate of 5-9% with 7-17% of minimum computational cost
compared with the ID-based Multiple secret Key Management (IMKM).
The second research work deals with node cooperation to avoid
security issues like illogical node participation, with Hybrid Approach for
Node Cooperation based Clustering (HANCC) in MANET. The node
ii
cooperation among the nodes in MANET is improved by forecasting the
weightage of cooperativeness of each node in the network. The evaluation of
node cooperation weightage detects the illogical nodes participating in the
network. Performance evaluations are carried out. HANCC achieves the node
cooperativeness of 20-25% compared with the Fair, efficient and secure
cooperation incentive mechanism (FESCIM) and efficient clustering and
cluster head rotation scheme for wireless sensor networks (ERP-SCDS).
The third research work, secures communication by the
establishment of Efficient Node Cooperation and Security (ECNS)
mechanism overcomes the bottleneck of selfish nodes in MANET by creating
initial authentication among nodes through node cooperation. The ENCS
mechanism is capable of providing high security by avoiding the misbehavior
nodes. Performance evaluations are carried out and it achieves the security
level of 15-25% compared with the Value iteration algorithm (VIA) and
ODMR protocol using a high-throughput (ODMRP-HT).
The final research work deals with Routing Aware Packet
Reserving (RAPR) framework for end-to-end throughput maintenance. This
resolves the effective resource allocation. The mobile node selects the packet
which travels in shortest distance. RAPR estimates the security level of the
system by end-to-end routing by controlling and clogging. The performance
evaluations are carried out and system uses the resources effectively and
improves the end-to-end throughput by 15%.
iii
TABLE OF CONTENTS
ABSTRACT
i
1
INTRODUCTION
1
1.1
BACKGROUND
1
1.2
STATEMENT OF THE PROBLEM
3
1.3
SECURITY IN MANET
4
1.3.1 Attacks in MANET
6
1.3.1.1 Vulnerabilities of MANETs
7
1.3.1.2 Attackers
8
1.3.1.3 Active and passive attacks
9
1.3.1.4 Attacks at physical layer
11
1.3.1.5 Attacks at network layer
12
1.3.1.6 Attacks at transport layer
14
1.3.1.7 Attacks at data link layer
15
1.3.2 Denial of Service Attack
15
1.3.2.1 DoS attacks on the link layer
15
1.3.2.2 DoS attacks on the network layer
16
1.3.3 Channel Aware Detection
17
1.4
NODE CLUSTERING IN MANET
18
1.5
PURPOSE OF THE STUDY
21
1.6
ORGANIZATION OF CHAPTERS
23
2
LITERATURE REVIEW
26
2.1
INTRODUCTION
26
2.2
A SECURE KEY MODEL FOR
EFFICIENT NODE CLUSTERING
BASED ON REPUTATION AND RANKING
26
iv
2.3
A HYBRID APPROACH FOR NODE
CO-OPERATION BASED CLUSTERING
IN MANET
35
2.4
EFFICIENT NODE COOPERATION
AND SECURITY IN MANET USING
CLOSENESS A DEGREE OF SEPARATION
44
2.5
RESEARCH GAP
52
2.6
OBJECTIVES OF THE RESEARCH
54
2.7
CONTRIBUTIONS OF RESEARCH
55
3
SECURE KEY MODEL WITH REPUTATION
AND RANKING SYSTEM IN MANET
56
3.1
INTRODUCTION
56
3.2
SECURITY ISSUES IN MANET
57
3.2.1 Reputation Based Self
Re-Organized Node Clustering
60
3.2.1.1 Reputation management
65
3.2.1.2 Reputation broadcast
66
3.2.1.3 Reputation detects, filter,
transform and localize
66
3.2.1.4 Resolver
67
3.2.1.5 Route maintenance
67
3.3
SECURE KEY MODEL ON NODE
CLUSTERING USING REPUTATION
AND RANKING
69
3.4
NODE CLUSTERING BASED ON
REPUTATION AND RANKING
71
3.5
GROUP KE MANAGEMENT
74
3.5.1 Need for Group Key Management
74
3.6
PSEUDO CODE FOR SECURE KEY MODEL
76
3.7
EXPERIMENTAL EVALUATION
78
v
3.8
RESULTS AND DISCUSSION
79
3.8.1 Malicious Node Detection Efficiency
79
3.8.2 Node Reputation
80
3.8.3 Performance Rate
82
3.8.4 Computational Cost
84
3.9
SUMMARY
86
4
A HYBRID APPROACH FOR NODE
CO-OPERATION BASED CLUSTERING
IN MANET
87
4.1
INTRODUCTION
87
4.2
NEED FOR NODE CO-OPERATION
BASED CLUSTERING IN MANET
88
4.2.1 Challenges of Node
Cooperation in MANET
91
4.2.1.1 Mobility management
91
4.2.1.2 Power control and bandwidth
allocation
92
4.2.1.3 Privacy and security
93
4.3
HYBRIDIZATION OF NODE
CO-OPERATION BASED CLUSTERING
95
4.3.1 Evaluation of Weightage of
Node Cooperativeness
96
4.3.2 Process of Self-Organization and Node
Clustering Based on Cooperativeness
100
4.3.2.1 Local activity rules to attain
universal characteristics
100
4.3.2.2 Development of understandable
coordination
101
vi
4.3.2.3 Reduction of long-lived
circumstances information
101
4.3.2.4 Protocols capable of any
topology change
101
4.3.3 Clustering Based on Weightage and
Self Organization of Nodes
102
4.3.4 Node Clustering Using HANCC
103
4.3.5 Algorithm for HANCC
107
4.4
EXPERIMENTAL EVALUATION
109
4.5
RESULTS AND DISCUSSIONS
109
4.5.1 Node Cooperativeness
110
4.5.2 Clustering Energy Dissipation
112
4.5.3 Network Lifetime
113
4.6
SUMMARY
115
5
EFFICIENT NODE COOPERATION
AND SECURITY IN MANET USING
CLOSENESS TECHNIQUE
116
5.1
INTRODUCTION
116
5.2
VULNERABILITIES OF THE
MOBILE AD HOC NETWORKS
117
5.2.1 Lack of Secure Boundaries
118
5.2.2 Threats from Jeopardize
Nodes Inside the Network
119
5.2.3 Lack of Centralized Management Facility 120
5.2.4 Restricted Power Supply
122
5.2.5 Scalability
123
5.3
ACHIEVING COOPERATION AMONG
NODES
126
vii
5.4
ENHANCING SECURITY OVER MANET
128
5.5
EFFICIENT NODE COOPERATION
AND SECURITY IN MANET USING
CLOSENESS TECHNIQUE
129
5.5.1 Algorithmic Flow of ENCS Mechanism
132
5.6
PERFORMANCE EVALUATION
134
5.7
RESULTS AND DISCUSSION
135
5.7.1 Average Information Leakage
136
5.7.2 Packet Transmission Efficiency
137
5.7.3 Security Level
140
5.7.4 Average Cost
141
5.8
SUMMARY
143
6
PACKET RESERVING AND CLOGGING
CONTROL VIA ROUTING AWARE PACKET
RESERVING FRAMEWORK IN MANET
144
6.1
INTRODUCTION
144
6.2
KEY IDEA BEHIND ROUTING AWARE
PACKET RESERVING FRAMEWORK IN
MANET
148
6.2.1 Packet Reserving on Mobile Nodes
149
6.2.2 Single RAPR Model with Multiple
Packet Flows
151
6.3
OVERVIEW OF ROUTING AWARE PACKET
RESERVING FRAMEWORK IN MANET
152
6.3.1 Algorithmic Description
of RAPR Framework
153
6.4
EXPERIMENTAL SETUP OF
RAPR FRAMEWORK
155
6.5
PERFORMANCE OF RAPR FRAMEWORK
156
viii
6.6
SUMMARY
165
7
CONCLUSION AND FUTURE WORK
166
7.1
CONCLUSIONS
166
7.2
FUTURE WORK
168
REFERENCES
170
1
CHAPTER 1
INTRODUCTION
1.1
BACKGROUND
A Mobile ad hoc network (MANET) is a system of wireless mobile nodes that
dynamically self-organize in random and momentary network topologies. A
MANET is group of wireless networks consisting of a number of mobile
nodes. Nodes in MANET connect and disconnect from the network
dynamically. There is no permanent set of infrastructure and centralized
administration in MANET. Nodes are organized and interconnected through
wireless interface. The self-motivated nature of such type networks makes it
extremely vulnerable to different link attacks. The essential needs for a
protected networking are secure protocols, ensuring that the privacy,
accessibility, authenticity and honesty of network. Many offered security
solutions for wired networks are unproductive and useless for MANET
environment. As the communication takes place in an open medium, the
MANET communication is more susceptible to security attacks. In the
occurrence of security protocol, the effects against different attacks are
reduced. Therefore, the success of MANET communication greatly depends
on the relationship of the participating mobile nodes.
A MANET is described by deficient in infrastructure, lack of
centralized administration, recurrent mobility of nodes, network separation
2
and wireless links. These characters indictate that the usual wire line security
solutions are not directly applicable in MANET. Certainly, the properties
represent that the establishment of a public key infrastructure, is an unwieldy
task in such networks.
Propagation of packet to the entire network is an essential process
and includes broad applications in the MANET. The basic approach for
transition is blind flooding, in which each node is compelled to retransmit the
data whenever it obtains a packet for the first time. Blind flooding creates
many repeated broadcasting. These repeated transmissions cause a serious
trouble, referred as the transmission storm difficulty in which repeated
packets leads to communication congestion and contention.
A MANET is a completely on-the-fly network utilized to maintain
the idea of any time and any place transmission. MANET is an infrastructure-
less network with a set of wireless mobile hosts to structure a temporary
network without the support of any wired base stations. Each mobile node in
such a network functions not only as a node but also as a router. The intrinsic
restrictions of the MANET, such as insufficient resources and dynamic
topologies, need a suitable routing protocol. The protocol design for such an
environment requires straightforward, well-organized and robust structure.
The methods that broadcast packets in the MANET are relatively unusual
from that in the wired network, since a node transmits a packet, then all its
neighbors receive that packet under the promiscuous receive mode.
Recent work on providing security comprises a fully-distributed
Identification based Multiple secret Keys Management scheme (IMKM)
facilitating an efficient key method. It involves an interaction of ID-based
multiple secret and threshold cryptography. IMKM eliminates public key
distribution using certificate authentication. In addition IMKM provides
efficient key update and key revocation schemes. The nodes are clustered and
3
needs a Cluster Heads (CHs) to participate in key construction. The updated
cluster head shares the keys on establishing a threshold sharing of the master
secret key. Multiple secret keys scheme is able to withstand cryptanalysis and
periodically updates share keys of CHs with a predefined time interval. The
scheme does not require the exchange or signing of any additional messages.
At last, the method supports improved performance by reducing computation
and communication overheads. Efficient group key agreement provides
authentication without authenticating signatures and needs only one round.
1.2
STATEMENT OF THE PROBLEM
Ensuring secure communication in ad hoc network is extremely challenging
due to dynamic nature of the ad hoc network. The difficulty in MANET
communication is that the data broadcast is insecure due to lack of centralized
management. MANET is a division of ad hoc network working in wireless
infrastructure-less environment in the way of self-configuring communication
with the mobile devices. Each node or mobile device in a MANET is
independent to each other moving without self-control in any route resulting
in insecure communication. The independent behavior of node flow changed
the interaction link to other devices or nodes frequently. The absence of fixed
infrastructure in shared wireless medium results in node mobility and limited
resources of mobile devices. Due to the limited resource in communication,
the bandwidth is restricted. Additionally, error-prone communication links by
key management is difficult to implement in ad hoc networks.
The information is broadcast with a secure private key. Group Key
Agreement (GKA) protocol is the common protocol generally used in secure
transmission. GKA protocol permits two or more parties to agree on common
group key and exchange information over insecure channel. GKA provides
mutual key authentication among parties involved in the communication.
4
The limitation of GKA protocol for transmission is its overcoming
nature with the establishment of Authenticated Group Key Agreement
(AGKAP) protocol applications in cooperative, distributed and self organized
ad hoc networks. Design of secure and well-organized protocol for group key
agreement gains much concentration as an important research area. The group
key management protocol is also less efficient in the transmission of packets.
Major researches are needed to be handled in the key management issues
for securing MANET using clusters.
1.3
SECURITY IN MANET
Wireless Mesh Networks (WMNs) are noted as a standard choice for Internet
Service Providers (ISPs) in broadcasting the information over wireless access.
The WMNs are estimated to integrate the characteristics of self-organization,
self-healing, and self-configuration for high consistency and scalability. In
addition to the numerous advantages, the WMNs require security guarantees
due to its open medium, shared architecture and inconsistent topology.
A MANET is an independent set of mobile nodes which
communicate over moderately bandwidth restricted wireless links. MANET
varies from predictable wireless networks, such as cellular networks and IEEE
802.11, an infrastructure mode networks as self-organizing. The infrastructure
networks are self-containing i.e., the nodes communicates openly with each
other lacking reliance on centralized infrastructures such as base stations.
Additionally, MANET is self organizing and adaptive in the way of
structuring and de-structuring on-the-fly without the need for any system
administration.
These exclusive characters make MANET popular for situations
which will need fast network use, such as search and rescue operations. The
decentralized property of MANET, specifically the deficiency of centralized
5
entities offers a better application work nature. The avoidance of single point
of disappointments makes these network models also perfect for military and
commercial applications that require high level of robustness. But some
challenging security issues are to be concentrated before MANET is set for
extensive commercial or military deployment.
One of the major security issues in MANET is trust management.
Trust is usually recognized and handled in wired and other wireless networks
using centralized entities like Centralized Authority (CA) or Key Distribution
Center (KDC). The lack of centralized entities in MANET creates trust
management security issue as a challenging task. The inaccessibility of trusted
authorities also builds crisis to achieve essential functions such as the
revocation of distribution centre. Another attractive MANET security
difficulty is the issue of false accusation in the existence of malicious nodes.
The false accusation tries to prove the valid node as malicious node resulting
in removal of legitimate node from the network. The malicious node causes
several communication difficulties such as gap of opportunity problem.
In MANET, certification systems play a vital role to attain network
security. Controlling the issue of certificate cancellation in wired network is
simple compared to the MANET. In wired network when the certificate of a
malicious node get canceled then the certificate authorities append the
information about the cancelled node into the Certificate Revocation Lists
(CRLs). Otherwise, they transmit the CRLs to each and every node present in
the network or either saves them on an open repositories. But the certificate
revocation is a difficult task in MANET and also this usual method of
certificate revocation is not valuable for MANET due to lack of centralized
repositories and trusted authorities. A method is required for MANET to
6
cancel the certificate of malicious nodes after detecting the first misbehavior
of nodes.
The wireless technology makes MANET more susceptible to
security attacks and due to this the established security methods does not offer
a novel solution to MANET. A new protocol needs to be urbanized to
overcome the disadvantage in the traditional security methods such as
distribution centre. Symmetric key cryptography method requires trusted third
party and vital repositories to preserve information about the node whose
certificate is get cancelled. But these standard security methods fail in
providing the preferred security in the case of wireless networks such as
MANET. In other words, the capacity of the standard security approaches is
only restricted to the wired networks and to some extent in the wireless
networks because the number of security problems is less in wireless
networks compared to wired networks.
1.3.1
Attacks in MANET
A MANET is an infrastructure-less category network containing number of
mobile nodes with wireless network boundary. In order to establish
communication among nodes, the nodes enthusiastically launch paths among
one another. The character and outline of such networks make it interesting to
several types of attackers. The section discusses different types of attacks on
various layers under protocol stack. Various types of attacker try different
approaches to drop off the network performance and throughput. The main
attention is on routing and security issues related with mobile ad hoc networks
which are necessary in order to offer secure communication. On the basis of
the character of attacks on communication, the attacks against MANET are
classified into two types, namely, active and passive attacks. Also the
7
attackers against a network are classified into two groups namely, insider and
outsider. An outside attacker is not a legal user of the network but an insider
attacker is a certified node and a component of the routing mechanism on
MANET.
1.3.1.1
Vulnerabilities of MANETs
A MANET is vulnerable to different attacks not only from exterior but also
from within the network itself. Ad hoc networks are essentially subjected to
several vulnerability issues as follows:
Dynamic Topology: In MANET, nodes connect and disconnect from the
network dynamically and travel independently. Due to such type character,
there is no permanent set of topology mechanism in MANET. The node with
insufficient physical security becomes malicious node and minimizes the
network performance.
Wireless Links: As the nodes in such networks are interrelated through
wireless interface that makes it extremely vulnerable to link attacks. The
bandwidths of wireless networks are not as much of wired networks and
offers less bandwidth. The less restricted bandwidth draws many attackers to
stop normal communication among nodes.
Cooperativeness: In MANET, all routing protocols believe that nodes offer
secure communication. But some nodes become malicious nodes which
interrupt the network process by modifying routing information.
Lack of clear line of defense: There is no clear line of protection mechanism
available in the MANET. Attacks come from any directions. Attackers attack
the network either within the network or outside the network.
8
Limited resources: The MANET includes various collections of devices such
as laptops, computers, and mobile phones and so on. All of such devices
consist of different storage capacity, processing speed and computational cost.
This attracts the attackers to concentrate on new attacks.
1.3.1.2
Attackers
There are various types of attackers present in MANET. They attempt to
minimize the performance of network. The various attackers are categorized
as shown in Figure 1.1.
Figure 1.1 Classifications of Attackers
Figure 1.1 describes the classification of attackers in MANET into five major
types in two levels. The first level of attack arises on the fundamental means
of the ad hoc network such as routing like emission, location and quantity.
The second level of attacks attempts to break the security methods engaged in
the network such as motivations and mobility.
Attackers
Emission
Location
Quantity
Motivations
Mobility
· Active
· Passive
· Insider
· Outsider
· Simple
· Multiple
· Confidentiality
· Integrity
· Selfishness
· Privacy
· Unauthorized
access
· Denial of
service
· Fixed
· Mobile
9
1.3.1.3
Active and passive attacks
A MANET is more vulnerable to passive attacks. A passive attack does not
change the data transmitted within the network. But it comprises the
unauthorized listening to the network traffic or gathers data from it. Passive
attacker does not interrupt the process of a routing protocol but tries to find
the significant data from routed traffic.
Identification of such type of attacks is complex since the process
of network itself does not get affected. In order to conquer this type of attacks
dominant encryption algorithms are required to encrypt the data being
broadcast. The attack against the MANET is kept on increasing due to its
open medium and independent nature. In addition to passive attack, another
attack that plays against secure communication is active attack.
Active attacks are very harsh attacks on the network that stop
message flow between the nodes. But active attacks are of inside or outside
type. Active outside attacks are handled by outside basis that do not belong to
the network. Inside attacks are from malicious nodes which are part of the
network. Internal attacks are more rigorous and inflexible to detect than
external attacks. These attacks make unauthorized access to network that
supports the attacker to make changes such as alteration of packets, denial of
service and congestion.
10
Figure 1.2 Active and Passive Attacks in MANET
Figure 1.2 describes the active and passive attacks in MANET. The passive
attacker listens to the data broadcasted in the network. The active attacker acts
as transmitter and changes the data transmitted to the receiver. Detecting the
active attacks is more difficult compared to passive attacks. The nature of
MANET formulates them vulnerable to many new attacks. The attacks in
different layers of the network protocol stack are described in
Table 1.1.
Active
Attacker
Passive
Attacker
Listens
to Data
Transmitter
Acts as
Transmitter
Active
Attacker
Changes Data
Receiver
MANET
11
Table 1.1 Attacks on Protocols
Layer
Types of Attacks
Application Malicious code, Data corruption, Viruses and Worms
Transport
Session hijacking attack and SYN flooding attack
Network
Blackhole, Wormhole, Grayhole, Link spoofing, Rushing
attack, Replay attack, Sybil attack, Resource consumption
attack and Link withholding attack
Data Link
Selfish Misbehavior, Malicious Behavior, Traffic Analysis
Physical
Eavesdropping, Jamming and Active interference
The attacks on physical layer are hardware related and require support from
hardware origins. These attacks are straightforward to perform as compared to
other attacks. They do not need the entire knowledge of technology. Some of
the attacks detected at physical layer comprise eavesdropping, intrusion and
congestion.
1.3.1.4
Attacks at physical layer
Eavesdropping is an interception and interpretation of messages and
discussions by unintended receivers. As the transmission takes place on
wireless medium, the message is easily interrupted with receiver tuned to
appropriate frequency. The major intention of such attacks is to take the
confidential information that is kept covered during the transmission. The
information includes private key, public key, place or passwords of the nodes.
Classified data is eavesdropped by tapping transmission paths, and wireless
links are easier to trace.
12
Figure 1.3 Attacks on Communication between Source and Destination
Figure 1.3 describes the attacker attacking on transmission between source
and destination. Jamming is a special class of DoS attacks which are initiated
by malicious node after determining the frequency of communication. An
active interference is a denial of service attack which stops the wireless
communication channel or alters communications. Old messages are repeated
to reinitiate the data information.
1.3.1.5
Attacks at network layer
The network layer protocols allow the MANET nodes to be linked with
another through hop-by-hop. In MANET, every distinct node takes choice of
route to transmit the packet. So, it is very simple for malicious node to attack
on such network. The essential thought behind network layer attacks is to add
itself in the active path from source to destination or to take up network
traffic. In such attacks, the attackers build routing loops to form harsh
congestion. Various type of attacks are detected which are initiated by
malicious node. Figure 1.4 describes the routing attacks by the malicious
nodes.
Original
Connection
Source
Attacker
Destination
13
Figure 1.4 Routing Attack by Malicious Nodes
The malicious node X takes main data by locating itself between source A and
destination D as shown in Figure 1.4. X also distracts the data packets
swapped between A and D, resulting in significant end to end delay between
A and D. The malicious node X interrupts the route path discovery process by
constructing routing loops and overflow of routing tables.
In Blackhole Attack, malicious node maintains finest path to the
node packet to be interrupted. On getting the request the malicious node
forwards a false respond with extremely short route. In wormhole attack,
malicious node obtains data packet at one location in the network and
channels them to another malicious node. Rushing Attacks are essentially
against the on-demand routing protocols. These types of attacks threaten the
path discovery process.
A
B
E
F
D
C
X
14
1.3.1.6
Attacks at transport layer
Attacker in Session Hijacking acquires the benefit to develop the insecure
session after its initial setup. In this attack, the attacker spoofs the injured
party node's IP address, discovers the right sequence number i.e. estimated by
the target and then initiates various DoS attacks. In Session Hijacking, the
malicious node attempts to gather secure data like passwords, secret keys,
logon names and other information from nodes. Session Hijacking attacks are
also known as address attack and affect the transmission protocol like
Transmission Control Protocol (TCP).
Figure 1.5 Session Hijacking
Figure 1.5 describes the Session Hijacking. The TCP- acknowledgement (ACK)
tempest trouble arises when malicious node initiates a TCP session hijacking
attack. The attacker X adds session data, and node 1 forwards
acknowledgement packet to node 2. Packet does not hold any sequence
number that node 2 is expecting. The attack is seen when node 2 receives the
packet and tries to resynchronize the TCP session with node 1. This process is
X
Injects Data in Session
Node 1
ACK
Node 2
15
repeated over and over that leads to ACK tempest. Application layer protocols
are also vulnerable to many DoS attacks.
1.3.1.7
Attacks at data link layer
Selfish
misbehavior
of
node
straightforwardly
changes
the
self-performance of nodes and does not obstruct the action of the network.
The main objective of malicious node is to interrupt regular operation of
routing protocol. The collision of such attack is enlarged when the
communication takes place between adjacent nodes.
The significant attack at data link layer is the DoS attack. These
types of threats created a malicious action with the support of node
co-operation that forms harsh security risks. In the presence of node cooperation,
it is very complex to identify the cooperation routing. The cooperation route
emerges like a usual route but guides to rigorous problems. For example, node
cooperation involves in the communication but drops some packets resulting in
degradation of service quality being offered by network. A detailed overview of
DoS attack is described in the following section.
1.3.2
Denial of Service Attack
A DoS attack is an occurrence that reduces or removes a network's power to
perform its expected function. The hardware failures, software errors,
resource collapses, environmental constraints, or any complex
communications between these factors cause DoS. The DoS attacks on the
link layer and network layer are summarized below:
1.3.2.1
DoS attacks on the link layer
IEEE 802.11 Medium Access Control (MAC) protocol is used as the link
layer protocol for MANET. IEEE 802.11 MAC is vulnerable to DoS attacks
16
and utilizes its binary exponential back-off scheme. Because a successful
broadcast results in a smaller conflict window, a constantly broadcasting node
always confine the channel and causes other nodes to back off continually. A
customized back-off scheme solves DoS attack by offering the back-off timer
from the receiver end. Additionally, the Network Allocation Vector (NAV)
field in the Request to Send/Clear to Send (RTS/CTS) frames represents
another vulnerability to DoS attacks. Since a malicious node is conscious of
the time interval of the current transmission in its neighborhood, it forwards
few bits to interrupt the current link-layer frames with an insignificant energy
cost.
1.3.2.2
DoS attacks on the network layer
DoS attacks on network layer normally categorized into three, namely,
resource deficiency, routing interference and forwarding denial.
In a resource deficiency attack, malicious nodes add additional
control or data packets into the network. For example, a malicious node keeps
forwarding different messages to its adjacent node. Since the sequence
numbers or fake target address is changed each time, an attacker's neighbors
are unable to distinguish the messages as fake ones or new requests. If the
malicious node forwards these fake messages at a high speed, its neighbors
spends much resources, like bandwidth, CPU sequences and battery power, to
face fake messages.
A typical example of this attack is Blackhole in which an attacker
initiating the Blackhole Attack could direct all packets to some destination
and then discard them. Another type of routing interruption attack is so called
wormhole. To begin the wormhole attack, two malicious nodes N1 and N2
cooperate with each other via a private network connection, such that N1
forwards the packets received from other nodes directly to N2 through the
17
wormhole. N2 rebroadcast the received packets to another area of the
network.
1.3.3
Channel Aware Detection
The WMN is a multi-hop network which depends on mesh routers to send the
packets to the destination. A successful association among routers is the basis
for a trustworthy network. Cryptography solutions are utilized to defend the
mesh routers from most of the routing protocol attacks like selective
forwarding, Gray hole, Blackhole and Wormhole Attacks. But, if the routers
are liberal, the attacker is allowed to access the public/private keys of the
liberal routers and then crack through the cryptographic system. Therefore, to
attain absolute security in a network, it is ideal to use cryptographic solutions
as a first line of protection and non-cryptographic solutions as a second line of
protection.
Most of the existing studies on selective forwarding attacks
concentrate on attack detection under the statement of an error-free wireless
channel. An additional realistic and demand scenario that packet falling is due
to Grayhole Attacks or normal loss events such as medium access crash or
worst channel quality. Specifically, a Channel Aware Detection (CAD)
algorithm is developed to effectively detecting the selective forwarding
attackers by sorting out the standard channel losses.
The CAD algorithm is based on two measures, namely, channel
evaluation and traffic monitoring. The measure of channel evaluation is to
approximate the normal loss rate due to worst channel quality or medium
access crash. The measure of traffic monitoring is to monitor the definite loss
rate. CAD involves four-fold contributions:
18
Channel evaluation is incorporated with traffic monitoring to attain
channel-aware detection of Grayhole Attack. The Grayhole detection
effectively detects selective forwarding misbehavior unseen in the usual loss
events due to worst channel quality or medium access crash.
In CAD, upstream and downstream traffic monitoring is integrated
to obtain a flexible detection method. In addition to Grayhole Attack, the
CAD also identifies restricted transmit-power attack, on-off attack and
dreadful opening attack.
The CAD algorithm is inefficient when multiple malicious nodes
act in collision. The CAD is unable to provide useful information about the
authenticated and unauthenticated nodes in the network. The more
unauthenticated nodes involved in the communication, instead of
authenticated node, lead to unsecure communication. Additionally, the
unauthenticated nodes are not prevented to involve in the network
communication.
1.4
NODE CLUSTERING IN MANET
Cluster-based routing is a solution to address node's diversity and to control
the quantity of routing information that broadcasts inside the network. The
purpose of clustering is to collect the network nodes into a number of
coinciding clusters. Clustering makes promising ranked routing in which
routes are traced between clusters instead of nodes. This increases the routes
duration, thus reducing the amount of routing control overhead. In the cluster
formation, the node coordinating the cluster activities is called as cluster head
(CH).
For a clustered network, the network is grouped into clusters with
one cluster head per cluster. Essentially, the clustered network transfers a
19
thick network to a thin one that involves cluster heads and some gateways.
The broadcast protocol uses a separation of nodes, called forward node set, to
communicate a broadcast packet in a clustered network. Only a cluster head
determines its forward node set to envelop other cluster heads within its
neighborhood and within the exposure area. A non-cluster head node
immediately communicates the broadcast packet if it is chosen as a forward
node. The forward node set is determined by cluster heads. All the cluster
heads in the network are connected to each other. Consequently, a broadcast
packet is delivered to the total network ultimately. Cluster heads are chosen
through a selection process. A cluster head straightforwardly connects to all
the nodes in the cluster. Other members in the cluster are non-cluster head
nodes. The clustered network is produced by the lowest-ID cluster algorithm.
Figure 1.6 Sample Clustered Network
Figure 1.6 shows the effect of applying clustering in a network with 10 nodes.
The cluster formation involves the connection of nodes into clusters groups.
The node 2 acts as attacker in interrupting the communication. Other nodes
Cluster 4
Cluster 1
Cluster 2
Cluster 3
1
4
2
7
8
9
10
3
5
6
Attacker
Cluster Gateway
Member/Cluster Node
20
are protected to deliver the packets. The pure clustered network does not
sustain position of maintenance. But it will be possible to localize if a little
different cluster structure strategy is applied. Once a cluster is shaped, a non-
cluster head node, which are newly entered in the cluster challenges the
current cluster head.
Figure 1.7 describes the cluster graph and adjacent cluster graph of
the sample network as shown in Figure 1.6. If a cluster head travels into an
existing cluster, the cluster head that has the higher ID will stop its role of
cluster head. If a cluster head travels out of a cluster, the left non-cluster head
nodes in this cluster will compute their new clusters. A node that has cluster
head adjacent obtains the neighboring cluster head with the lowest ID as its
new cluster head and connects in that cluster. For nodes with no cluster head
adjacent, the cluster formation procedure is applied among those nodes to
structure new clusters. Thus, the clusters mobility is adaptive and alters a
cluster, controlled in a limited area. Thus, transmission of packet in a
clustered MANET using the forward node set is able to efficiently convey the
packet.
Figure 1.7 explicates the cluster graph formed in a cluster. Each
cluster has its own nodes with its edges. In this the vertices refers the nodes
and the edges refers the connection between the nodes. The cluster graph in
Figure 1.7(a) is with good quality and dense, because all the nodes are close
to each other. The cluster graph in Figure 1.7(b) is almost similar to Figure
1.7(a) with same number of nodes and edges; whereas most of the vertices are
outside the cluster. The cluster graph in Figure 1.7(c) has very few
connections with other nodes but lacks in internal density and hence it is not a
good cluster.
21
Figure 1.7 Cluster Graph and Neighbor Cluster Graph
1.5
PURPOSE OF THE STUDY
Security is the main challenge faced in protecting wireless communications
especially in MANET. The security test reveals that many techniques provide
strong privacy protection but wormhole attacks are difficult to prevent. These
existing techniques support a secure shared key based on-demand for packet
data but offers an unsecure communication.
In MANET topology, crisis occurs mainly due to the attacks that
affect the network. A dynamic method is required to employ topology control
algorithms in MANET to avoid topology problems. Network topology among
MANET varies due to mobility and thus, it unable to maintain the network
connectivity.
The cryptographic protocols concentrates on providing secure
communication but also affected by diverse attacks. The attacks present in the
networks and the root of that attack is also avoided. But the drawback is that
the attacks on encrypted protocol are unavoidable. For the attacks like
Blackhole, Grayhole, Wormhole and Rushing attack, no algorithms is
launched still now and above all attack problems are still faced by MANET.
The possible solution to attacks is by performing node clustering
and deriving cluster heads. Cluster heads are combined and used to avoid
(b)
(c)
(a)
22
attacks but at the same time a key security is required to solve the attacks. The
data transmission through the cluster nodes is efficient regarding energy but
the congestion at the cluster heads raises. Key security along with clustering is
necessary to avoid loss of packets or prevention from attacks.
A fully-distributed identification based multiple secrets key
management for a secure communication using a protocol of group key
management is proposed. But the disadvantages in IMKM are bandwidth
usage is limited for a secure communication and consumes several rounds to
recognize the secure channel. But the main drawback is that the
communication between the nodes and the clustering processes are inefficient.
Misbehavior of nodes also causes problems in MANET. Routing
misbehavior is also a difficult problem in providing security during
transmission as the nodes drops the packets before transmitting. The packet
dropping problem is avoided by using reputation based method which mainly
avoid when an attacker resets a poor reputation by rejoining the system with a
new identity
(
whitewashing attacks). The existence of delay problems keeps
on increasing due to addition of nodes resulting in inefficient packet delivery
to destination node in given period of time.
Nodes in the networks are scattered reducing the efficiency in
communication. Cluster-based routing is a resolution to solve nodes diversity
and to control the amount of routing information that propagates inside the
MANET. The nodes in the network are clustered for secure ad hoc
communication. A secure key model is provided for ad hoc network with
efficient node clustering based on reputation and ranking model. An
improvement in the reliability of the network is required to provide a fast
transmission. The reputation technique is necessary to find the
unauthenticated nodes involved in the communication.
23
In addition, the reputation and ranking models are employed to
detect the misbehavior, and selfish nodes within the network communication.
Determination of unauthenticated nodes is essential to avoid the adversary
acts performing in MANET. The nodes co-operation is necessary to perform
clustering and in recognizing cluster head. The cluster head formation helps in
ranking.
A secure transmission between the source and destination plays a
vital role in the MANET communication. To provide the security in
communication, a key is incorporated with the information. To decide the
secret keys in authenticating group, an authenticated group key agreement
protocol is essential. There are several security issues generated by the selfish
nodes in MANET and to improve the cooperativeness among the mobile
nodes in MANET, it is necessary to ensure the node cooperation among the
nodes.
1.6
ORGANIZATION OF CHAPTERS
Chapter 1 provides a brief discussion on the security of MANET with
infrastructure-less network with different attacks on each layer of the network,
and clustering of nodes by providing efficient node cooperation between the
nodes. Additionally the performance and organization of nodes within the
network is deeply discussed.
Chapter 2 reviews the previous research work on MANET security,
infrastructure-less network that is relevant for the present research. In
addition, reviews related to node clustering in infrastructure-less environment
for secure communication, a secure key model for efficient node clustering
based on reputation and ranking, a hybrid approaches for node co-operation
based clustering in MANET using the closeness between the nodes are also
interpreted from the research works.
24
Chapter 3 provides an overview of various security related issues
while constructing the network. A secure key model (SKM) with reputation
and ranking system is proposed as a first research work. This model
dynamically evaluates the authentication of neighboring nodes and selects the
trusty nodes in the cluster to support efficient transmission of data. The nodes
are clustered based on the reputation table. The model provides secure
successful communication between the nodes in the network without any loss
of data. The design of node clustering proves the usefulness and feasibility of
secure key model approach.
Chapter 4 presents a detailed description of a secure
communication framework between the nodes in the network without any loss
of data and the needs for node cooperation based clustering in MANET. The
second research work namely hybrid approach for node cooperation based
clustering (HANCC) is efficiently designed for enhancing a secure
communication over MANET by improving the node cooperation among
nodes by monitoring the behavior and activity of nodes along with the
weightage of node cooperation. The work evaluates the performance results of
the secure communication framework for hybrid approach using node
cooperation based clustering.
Chapter 5 describes a mechanism to avoid the misbehavior nodes
from replacing the security associations with unidentified nodes. The third
research work ENCS is designed for enhancing cooperation of nodes and
secures communication by adopting the closeness technique.
Chapter 6 focuses on end-to-end route awareness and represents the
end-to-end route quality in terms of path lifetimes. The fourth research work
Routing Aware Packet Reserving (RAPR) framework is developed in
MANET that takes into account both the clogging state and the end-to-end
throughput maintenance. RAPR is complimentary system to packet reserve
25
that utilizes only local routing information. RAPR framework local routing
information contains the node clustering, cooperation and high security level,
which provides the maximal throughput among the contending flows. Finally,
Chapter 7 provides the concluding remarks and suggestions for future work.
26
CHAPTER 2
LITERATURE REVIEW
2.1
INTRODUCTION
An ad hoc network is referred as wireless network and provides a wireless
network communication without using infrastructure mechanism.
Ad hoc is a Latin word and has the meaning "for this purpose". A mobile ad
hoc network is a part of ad hoc network defined as a self-configuring
infrastructure-less network, each communicated by wireless links with the
help of mobile devices. The nodes in the ad hoc network are independent of
each other and free to move. Since nodes are independent, packet losses are
found to occur while transferring the packets from source to destination. Due
to the dynamic nature of network and node mobility, the unsecure ad hoc
network loses packets. Each node or device in a MANET can move freely in
any direction. Always some changes occur in the communication links.
2.2
A SECURE KEY MODEL FOR EFFICIENT NODE
CLUSTERING BASED ON REPUTATION AND RANKING
Privacy plays a major role in forwarding a packet. Privacy-preserve routing is
vital for ad hoc networks that entail stronger privacy protection. Data packets
and control packets are still linkable and distinguishable in the old schemes
but they do not tender inclusive unlinkability or unobservability property.
Hence, Wan et al (2012) presented capable privacy requirements concerning
27
privacy-preserving routing in mobile ad hoc networks. An Unobservable
Routing Protocol (USOR) was designed and it provided unlinkability and
content unobservability to almost all forms of packets. USOR is capable,
because it formed a new merging of cluster signature and ID-based encryption
for route discovery. The security examination exhibits that USOR offers
strong privacy protection as well as it provided security against attacks which
are caused because of node compromise. But wormhole attacks were not
prevented using USOR.
Thus, to provide protection against wormhole attacks a security
protocol is to be proposed which is basically a difficult task. For
communication anonymity, (Sankara & Bhagyaveni 2008) used Secure On
demand position based private routing protocol (SO2P). It provided security
in mobile ad hoc networks, by means of developing a cryptography algorithm
to prevent message hacking. Global Positioning System (GPS) provided the
position information of nodes for which privacy was more important. GPS
obtained the position information of nodes in the network by a terminal node
which behaved like a server. All the information about the terminal nodes
present in ad hoc network was recorded by the server with the process of GPS
system. The position information was prohibited from internal and external
hackers by SO2P. It was performed using a secure routing algorithm in
MANET to identify optimal path to reach the destination. Because of security
mechanisms, SO2P faced some trouble in routing performance such as high
packet loss ratio, less throughput, more time in end to end connection delay.
Since SO2P was used under on demand basis, the secure communication
might not be an efficient one.
MANET is to be secure from attacks and must be dynamic in
providing security (Pankaj & Rajender 2009). Advancement in algorithms and
protocols were mandatory for a secure ad hoc network because of heavy rise
28
in security threats. Attackers or malicious nodes affect communication
process and it could be avoided using Encrypted Dynamic Source Routing
(EDSR). EDSR was also used to prevent numerous types of Denial-of-Service
attacks. If malicious agents were found on the pathway of forwarding a
packet, EDSR could drop the data packet or acknowledgment, thus preventing
malicious node not to get the right information. Data packets were delivered
to their destinations by EDSR but when different malicious agents were
presented, it drops the packet which was to be forwarded.
Nodes might also cause problems while packets were to be received
by a particular node. The nodes that were engaged in routing and forwarding
must have cooperation then only the MANET performs perfectly. Steps were
to be taken to make a system work properly even during the existence of
malicious nodes. The performances of MANET were greatly affected because
of selfish or malicious nodes. Such misbehavior nodes were to be managed.
Praveen Sam et al (2008) considered two factors, namely, direct factor and
distributed factor, called DUAL factor for the management.
As a development to reactive source-routing protocol, the DUAL
factor protocol was designed for MANET. DUAL factor holds the
components namely, the observer, the trust analyzer, the reputation scheme
and the route analyzer. These components were available in every node.
In DUAL factor each node traced the character of its next-hop
neighbors. If any event was found to be suspicious, data to be forwarded was
given to the reputation scheme. The node activity could be based on the
topology created among the network.
In MANET topology control was a challenging problem. K-edge
connected topology control algorithms were designed in order to build robust
topologies for mobile networks but it seemed to be insufficient. Normally in
29
networks, node moves at diverse speeds and hence, consistently applying k
values for localized topology control algorithms in any local graph was not
useful. Hence, a dynamic method was presented by Nishiyama et al (2012) to
use k-edge connected topology control algorithms in MANET. The proposed
dynamic method automatically decided the appropriate k value for all local
graphs based on local information. The trade-off among topology control and
reliability was compromised by using scalable topology control algorithms.
Network topology among MANET varied due to mobility and thus it could
not manage the network connectivity. The above mentioned was a well
known problem, but has yet to be solved. Topology problem arose mainly due
to the attacks that affected the network.
The great increase in network based applications had resulted in
numerous security leaks. The cryptographic protocols which provide secure
communication were also affected by diverse attacks. The network traffic and
host activities were monitored by Intrusion Detection Systems (IDSs) in order
to avoid impacts caused by unauthorized accesses and attacks. But the
traditional misuse-based and anomaly-based IDSs were vulnerable to attacks
that affected the encrypted protocols because they were always concerned of
payload contents inspection. An anomaly-based detection system with the
help of strategically distributed Monitoring Stubs (MSs) was presented by
Fadlullah et al (2010). The encrypted traffic was identified by MSs which
remove features for detecting the attacks. MSs also trace back the originating
network of the attack. The approach was called as DTRAB because it aimed
on both Detection and TRAceBack in the monitoring stubs level. But the
drawback was that attacks on encrypted protocols were not avoided.
MANET was a dynamic wireless network developed without any
preceding infrastructure and so each node could behave like a router and
MANET had no restriction (Khokar et al 2010). So it can be accessed by any
Details
- Pages
- Type of Edition
- Erstausgabe
- Year
- 2017
- ISBN (PDF)
- 9783960676850
- ISBN (Softcover)
- 9783960671855
- File size
- 5.5 MB
- Language
- English
- Publication date
- 2017 (November)
- Keywords
- IT Security Network security Secure key Re-Organized Node Clustering Node Co-Operation Mobile ad hoc network Wireless mobile host Temporary network Secure communication
