在無線隨意式網路中可抵抗蟲洞攻擊

國

立

交

通

大

學

資訊科學與工程研究所

碩

士

論

文

在無線隨意式網路中可抵抗蟲洞攻擊

之路由協定

An auto-resilient routing protocol against

wormhole attacks in Mobile Ad-hoc network

研 究 生：王偉民

指導教授：蔡文能 教授

i

在無線隨意式網路中可抵抗蟲洞攻擊之路由協定

An auto-resilient routing protocol against wormhole attacks in mobile

Ad-hoc network

研 究 生：王偉民 Student：Wei-Ming Wang

指導教授：蔡文能 Advisor：Wen-Nung Tsai

國 立 交 通 大 學

資訊科學與工程研究所

碩 士 論 文

Submitted to Institute of Computer Science and Engineering College of Computer Science

National Chiao Tung University in partial Fulfillment of the Requirements

for the Degree of Master

In

Computer Science May 2009

Hsinchu, Taiwan, Republic of China

ii

在無線隨意式網路中可抵抗蟲洞攻擊之路由協定

研究生： 王偉民 指導教授：蔡文能教授

國立交通大學

資訊科學與工程學系碩士班

摘要

iii

An auto-resilient routing protocol against wormhole attacks in

mobile Ad-hoc network

Student: Wei-Ming Wang Advisor: Wen-Nung Tsai

Department of Computer Science and Information Engineering

National Chiao-Tung University

Abstract

The Mobile Ad-hoc networks (MANETs) are self-configuring network and each node in

MANETs is free to move and has the ability to route packets. However, these characteristics

will give rise to Wormhole attack which will increase the influences of network attacks. In

recent years, the methods how to avoid wormhole attacks have become attractive research

issues. However, many previous works focus on observing the behavior of wormhole node to

solve the attack issues. We thought against wormhole attack utilizing the routing

characteristics of MANETs will has the better resulting of avoiding attacks. We researched the

routing protocol of MANETs, present the wormhole attacks using the weakness of routing

iv

誌

謝

v

Contents

Contents ... v

List of Figure ... viii

Chapter 1 Introduction ... 1

1.1 Research Motivation and Goals ... 2

1.2 Thesis Organization ... 2

Chapter 2 Background ... 4

2.1 Routing Protocols in mobile Ad-Hoc Networks... 5

2.1.1 Table-Driven Routing Protocols ... 6

2.1.2 Reactive Routing Protocols ... 7

2.1.3 Hybrid Routing Protocols ... 7

2.2 Security Issues in Mobile Ad-Hoc Networks ... 8

2.2.1 The Vulnerable Characteristic of Mobile Ad-Hoc Networks ... 8

2.2.2 Routing Security in MANETs... 10

2.3 Wormhole Attacks in Mobile Ad-Hoc Networks ... 11

2.3.1 Wormhole Attack ... 12

2.3.2 The effects of Wormhole Attacks ... 15

vi

3.1 Wormhole Attacks Prevention ... 18

3.1.1 Distance or Time Limiting Detection Approaches ... 19

3.1.2 Topology detection Approaches ... 20

3.1.3 Graph Theoretic and Geometric Approaches ... 21

3.1.4 Other Mechanisms and Protocols ... 22

3.2 Comparison of Main Mechanisms ... 23

Chapter 4 The Proposed Protocol: AR-AODV Routing Protocol ... 25

4.1 Observation of Wormhole Attacks on AODV ... 25

4.2 The Proposed Protocol: AR-AODV ... 29

4.2.1 Design Conception ... 30

4.2.2 Route Establishment ... 32

4.2.3 Route Maintenance... 33

Chapter 5 Simulation and Results ... 36

5.1 Simulation Environment ... 36

5.2 Simulation and Results ... 39

5.2.1 The impact of wormhole attack ... 39

5.2.2 Overhead of AR-AODV ... 41

Chapter 6 Conclusion and Future Work ... 46

vii

6.2 Future Work ... 46 Reference ... 48

viii

List of Figure

Figure 2-1 Classification of routing protocols in MANET ... 6

Figure 2-2 Wormhole attack ... 11

Figure 2-3 Wormhole attack against Routing ... 12

Figure 2-4 Illustration of wormhole attack ... 13

Figure 4-1 Routing request ... 27

Figure 4-2 wormhole attack on AODV ... 28

Figure 4-3 Path Record ... 29

Figure 4-4 Reverse RREQ ... 30

Figure 4-5 Illustration of route establishment ... 32

Figure 5-1 simple grid (10M*10M, 50 nodes) ... 38

Figure 5-2 large grid (50M*50M, 1000 nodes)... 39

Figure 5-3 impact rate (simple grid) ... 40

Figure 5-4 impact rate (large grid) ... 41

Figure 5-5 latency of RREQ packets (simple grid) ... 42

Figure 5-6 latency of RREQ packets (large grid)... 43

ix

1

Chapter 1 Introduction

In recent years, the MANET (mobile ad hoc network) technology has become a mature

technology and also it has been more and more popular. The MANET is a self-configuring

network of mobile devices connected by wireless links. Each device in a MANET is free to

move independently in any direction, and will therefore change its links to other devices

frequently. In comparison with traditional network environment that depends on some

infrastructure, mobile ad hoc network can communication each other without infrastructure.

However, some characteristics of mobile ad hoc network will increase their vulnerability

to attacks. Wireless links are inherently vulnerable to eavesdropping and message injection, as

well as jamming attacks. In these security issues, security routing in MANET is an especially

hard task to accomplish securely, robustly and efficiently. The MANET is constrained in

memory, computing power, and battery power in mobile devices. These restrictions will result

that routing protocol of MANET should be faced a tradeoff between security and resource

consumption.

In this theme, we conduct a study for the wormhole attack and we proposed a routing

2

attacks and it is very suitable for MANET environment due to it does not require too much

computing or any special hardware.

1.1 Research Motivation and Goals

In our research, we introduce the wormhole attack which has formed a serious threat in

wireless networks, especially against many ad hoc network routing protocols and

location-based wireless security systems. The wormhole attack is possible even if the attacker

has not compromised any hosts and even if all communication provides authenticity and

confidentiality. Also, most existing ad hoc network routing protocols, without some mechanism

to defend against the wormhole attack. In this theme, we design an effect approach to against

wormhole attacks and this approach can not only detect the wormhole attack but also auto

resilient from wormhole attack. Many of the existing methods have requirements of time

synchronization, GPS receiver, etc. However, our proposed protocol does not require special

hardware and only need few computing resources.

1.2 Thesis Organization

The rest of this thesis is organized as follows: In Chapter 2, we discuss the background

3

3, we discuss the related works and compare their advantages and disadvantages. In Chapter 4,

we present our Auto-resilient Routing against Wormhole Attacks in Ad Hoc Networks. In

Chapter 5, we discuss our evaluation methods and the results. Finally, we conclude this thesis in

4

Chapter 2 Background

In the section, we will discuss three topics about mobile Ad-Hoc networks include routing

protocol of MANETs, related security issues of MANETs and wormhole attacks. Unlike

traditional wired network or infrastructure-based wireless network, each node in mobile

Ad-Hoc network has routing capabilities; the prototype of mobile Ad-Hoc network composes

when a collection of mobile nodes link together and create a network by agreeing to route

messages for each other. There is no shared infrastructure in an Ad-Hoc network such as

centralized routers or defined administrative policy.

However, these feathers of ad hoc network routing protocol bring new security problems.

Unlike traditional wire network or infrastructure-based wireless network, each node in Ad-Hoc

network has routing capabilities and can be regarded as independent; it means each node in the

Ad-Hoc networks can arbitrarily attack other nodes by changing the routing rules. We will

introduction these security issues in section 2.2 and Wormhole Attack, the most intractable

5

2.1 Routing Protocols in mobile Ad-Hoc Networks

A mobile Ad-Hoc network is a collection of mobile nodes that are dynamically and

arbitrarily located in such a method that the interconnections between nodes are capable of

changing on a continual basis. In order to facilitate communication within the network, a

routing protocol is used to discover the path between nodes. An Ad-Hoc routing protocol is a

convention, or standard, that controls how nodes decide which way to route packets between

computing devices in a mobile ad-hoc network. The primary goal of such an ad hoc network

routing protocol is exact and efficient route establishment between a pair of nodes so that

messages may be delivered in a timely manner. Route construction should be done with a

minimum of overhead and bandwidth consumption.

In Ad-Hoc networks, nodes do not start out familiar with the topology of their networks;

instead, they have to discover it. The basic idea is that a new node may announce its presence

and should listen for announcements broadcast by its neighbors. Each node learns about nodes

nearby and how to reach them, and may announce that it, too, can reach them.

Classification of routing protocols in MANET’s can be done in many ways, but most of these are done depending on routing strategy and network structure [13][14]. According to the

6

Figure 2-1 Classification of routing protocols in MANET

2.1.1 Table-Driven Routing Protocols

These protocols have the same characteristics that they maintain the routing information

before it is required. For this characterize, the protocols also called as proactive protocols.

Every node in the network maintains routing information to other node in the network. Routes

information is usually kept in the routing tables and is periodically updated when the network

topology changes. Many of these routing protocols come from the link-state routing. There are

some differences between the protocols in this group depending on the routing information and

the routing table. These protocols are not suitable for larger networks, as they need to maintain

node entries for each and every node in the routing table of every node. This causes more

7

2.1.2 Reactive Routing Protocols

These protocols have the main characteristic that they don’t maintain routing information

or routing activity in advance when the network in initial step. If a node wants to send a packet

to another node then this protocol searches for the route in an on-demand manner and

establishes the connection in order to transmit and receive the packet. The route discovery

usually occurs by flooding the route request packets throughout the network.

2.1.3 Hybrid Routing Protocols

This type of protocols couples the benefit of proactive and of reactive routing like HSLS (Hazy

Sighted Link State routing protocol) [15], ZRP (Zone Routing Protocol) [16], etc. The routing

is initially established with some proactively prospected routes and then serves the demand

from additionally activated nodes through reactive flooding. The choice for one or the other

method requires predetermination for typical cases. The main disadvantages of such algorithms

are

1. Benefit should rely on number of nodes in network.

8

2.2 Security Issues in Mobile Ad-Hoc Networks

In recent years MANETs (mobile ad hoc networks) have received great attention due to

their capabilities of self-configuration and self-maintenance. While security issues have

become a primary interest in order to provide safe communication. Mobile Ad-Hoc networks

were proposed to support dynamic scenarios and situations where no wired network

infrastructure exists. In fact most Ad-Hoc routing protocols are collaborative by nature and rely

on implicit trust among the participating nodes to route packets through without any security or

information integrity guarantees. However, this is not always suitable especially when

information security and integrity is an important issue of the application

Although security issues have been an active research topic for a long time in wired

networks and there are a lot of research results. However, unlike the wired networks the

unique characteristics of MANETs present a new set of nontrivial challenges to security design,

such as open peer-to-peer network architecture, shared wireless medium, stringent resource

constraints and highly dynamic network topology.

2.2.1 The Vulnerable Characteristic of Mobile Ad-Hoc Networks

In general, the wireless MANET is especially defenseless due to its fundamental

9

cooperation, and constrained capability. In this section, we briefly list the following

vulnerable characteristics about MANET.

First, use of wireless links causes an Ad-Hoc network vulnerable to security attacks

include passive eavesdropping and active impersonation, message replay, and message

distortion. Eavesdropping might give an adversary access to secret information, violating

confidentiality. Active attacks might allow the adversary to delete messages, to inject erroneous

messages, to modify messages, and to impersonate a node, thus violating availability, integrity,

authentication, and non-repudiation.

Secondly, nodes, roaming in an unfriendly environment with relatively poor physical

protection, have non-negligible probability of being compromised. Therefore, we should not

only consider malicious attacks from outside a network, but also take into account the attacks

launched from within the network by inside nodes.

Thirdly, an Ad-Hoc network is dynamic because of frequent changes in both its topology

and its membership (i.e., nodes frequently join and leave the network). Trust relationship

among nodes also changes, for example, when certain nodes are detected as being

compromised. Unlike other wireless mobile networks, such as mobile IP nodes in an ad hoc

network may dynamically become affiliated with administrative domains.

Finally, an ad hoc network may consist of hundreds or even thousands of nodes. Security

10

2.2.2 Routing Security in MANETs

Routing protocols in MANETs are very different than traditional network, it has the

advantage of free configuration and high flexible. However, the features of routing protocols

in MANETs also causes many security attacks be launched by finding the loophole in routing

protocols in MANETs. Here we list the a few common attacks by utilizing the leak of routing

protocols of MANETs.

Black hole: In this attack, a malicious node uses the routing protocol to advertise itself

as having the shortest path to the node whose packets it wants to intercept. We provide a

detailed description herein.

Denial of service: The DoS attack results when the network bandwidth is hijacked by a

malicious node. It has many forms: the classic way is to flood any centralized resource so that

the network no longer operates correctly or crashes. For instance, a route request is generated

whenever a node has to send data to a particular destination. A malicious node might generate

frequent unnecessary route requests to make the network resources unavailable to other nodes.

Routing table overflow: The attacker attempts to create routes to nonexistent nodes. The

goal is to have enough routes so that creation of new routes is prevented or the

11

Information disclosure: The malicious node may leak confidential information to

unauthorized users in the network, such as routing or location information. In the end, the

attacker knows which nodes are situated on the target route.

2.3 Wormhole Attacks in Mobile Ad-Hoc Networks

Wormhole attack is a particularly severe control attack depends on the routing

functionality of wireless networks. It has been introduced in the context of Ad-Hoc networks.

During the attack, a malicious node captures packets from one location in the network, and

“tunnels” them to another malicious node at a distant point, which replays them locally. In fact, the wormhole attack is considered particularly insidious since it can be launched without

having access to any cryptographic keys or compromising any legitimate node in the network.

12

Figure 2-3 Wormhole attack against Routing

2.3.1 Wormhole Attack

Wormhole attacks happen when one wormhole node eavesdrops and records packets at

one location, and then tunnels the eavesdropped packets to a certain faraway collusive

wormhole node. After receiving the tunneled packets, the faraway collusive wormhole node

replays these packets (Figure 2-2). The tunnel between collusive wormhole nodes can be

established in various ways, such as direct wire connection, high power transmission, and

out-of-band hidden channel.

In a wormhole attack, an attacker forwards packets through a high quality out-of-band link

13

wormhole attack. The attacker replays packets received by X at node Y, and vice versa. If it

would normally take several hops for a packet to traverse from a location near X to a location

near Y, packets transmitted near X traveling through the wormhole will arrive at Y before

packets traveling through multiple hops in the network. The attacker can make A and B believe

they are neighbors by forwarding routing messages, and then selectively drop data messages to

disrupt communications between A and B.

Figure 2-4 Illustration of wormhole attack

For most routing protocols, the attack has impact on nodes beyond the wormhole

endpoints’ neighborhoods also. Node A will advertise a one-hop path to B so that C will direct packets towards B through A. For example, in on-demand routing protocols (DSR and AODV)

or secure on-demand routing protocols (SEAD, Ariadne, SRP), the wormhole attack can be

mounted by tunneling ROUTE REQUEST messages directly to nodes near the destination node.

Since the ROUTE REQUEST message is tunneled through high quality channel, it arrives

earlier than other requests. According to the protocol, other ROUTE REQUEST messages

14

routes from being discovered, and the wormhole will have full control of the route. The attacker

can discard all messages to create a denial-of-service attack, or more subtly, selectively discard

certain messages to alter the function of the network. An attacker with a suitable wormhole can

easily create a sinkhole that attracts (but does not forward) packets to many destinations. An

intelligent attacker may be able to selectively forward messages to enable other attacks.

To show how much damage a single wormhole can cause to routing, we simulated

randomly distributing nodes in a rectangular region and used the shortest path algorithm to find

the best route between any node pairs. If a wormhole is formed, some far away nodes will

appear to be neighbors and some node pairs will be able to find a “shorter” path through the

wormhole. Hence the route between them is disrupted by the wormhole. In simulation

experiments, a single wormhole with two randomly placed endpoints disrupts over 5% of all

network routes. A more intelligent attacker may be able to place wormhole endpoints at

particular locations. Strategically placed wormhole endpoints can disrupt nearly all

communications to or from a certain node and all other nodes in the network. In sensor network

applications, where most communications are directed from sensor nodes to a common base

station, wormhole attacks can be particularly devastating. If the base station is at the corner of

the network, a wormhole with one endpoint near the base station and the other endpoint one hop

away will be able to attract nearly all traffic from sensor nodes to the base station. If the base

15

quadrant of the network. Figure 2 shows the effectiveness of a wormhole in disrupting

communications from sensor nodes to a base station. One endpoint of the wormhole is within

one hop of the base station; the position of the second endpoint varies along the x axis. When

the base station is in a corner of the network, a wormhole with the second endpoint near the base

station can effectively disrupt all network communications. If the second endpoint is placed in

the opposite corner, approximately half of the nodes in the network will send messages for the

base station to the wormhole.

2.3.2 The effects of Wormhole Attacks

Wormhole attacks have very different features from other attacks: an adversary does not

have to breach the cryptography nor compromise any nodes for launching wormhole attacks,

because the adversary nodes need not decapsulate any frames or packets during attacks. What

an adversary needs to do is to setup collusive wormhole nodes in wireless networks, to capture

radio signals and to build tunnels between the collusive nodes. Then the wormhole nodes can

eavesdrop, capture and tunnel radio signals or steal private information that flows via collusive

wormhole nodes, although they have no identities or cryptography keys required in the network.

Additionally, since wormhole nodes replay signals at a place and can attract network traffic,

16

Hence, it is much more difficult to detect and prevent wormhole attacks. In fact, if the

wormhole nodes conduct no mal-behaviors in the network or are configured by network

administrators, wormhole tunnel may be a very pleasing feature. They may provide alternate

and faster routes, and even reduce the use of wireless bandwidth and save energy of mobile

nodes. But quite often the wormhole nodes may be laid down by malicious adversaries.

Wormhole attacks affect a network most significantly while nodes are establishing route to

another node. Wormhole attacks can make particular nodes in a network generate improper

routing tables for themselves. For example, if a wormhole attack is launched on a periodic

routing protocol, such as optimized link state routing protocol (OLSR), the collusive wormhole

nodes can let a regular node trust some other nodes are its neighbor nodes. In figure 1.2, node S

broadcast a Hello message periodically. If no wormhole nodes exist, only node A and B learn

that S is their neighbor nodes; however, if wormhole attacks exist, M1 can tunnel the Hello

message to M2, and M2 replays the Hello message. As a result, node W, X, Y and Z also believe

that S is their neighbor nodes

Besides periodic routing protocol, wormhole attacks can also cause great disruption on

on-deman routing protocol, such as DSR [9] and AODV [10]. Firstly, we suppose no wormhole

attacks happen in figure 1.3. We assume that S is attempting to establish a new route to W. S can

establish a good route to W by sending a route request (RREQ) through node A, C, D and W

17

-> A -> C -> D -> W. On the other hand, we assume that wormhole attacks exist, nodes.

It is worth noting that in the network, although both M1 and M2 physically exist, they virtually

vanish. M1 and M2 have no network identities but are repeaters in; i.e., none of the good nodes

in the network is able to aware of the existence of M1 and M2.

This is the major reason why wormhole attacks are difficult to deal with. As it is not able to

easily find out wormhole nodes in a network, a specific mechanism to prevent and detect

18

Chapter 3 Related Work

In this chapter, several research works related to wormhole attacks will be introduced.

These research works address the detecting mechanisms and prevention methods of Wormhole

Attack in mobile Ad-Hoc networks. For these methods defend against wormhole attack, it can

be classified into distance or time limiting detection approaches, false geometry or topology

detection approaches, and neighbor nodes monitoring approaches. We will discuss the theoretic

these three types of detecting approaches below. Moreover, we analyze and compare of these

methods.

3.1 Wormhole Attacks Prevention

Dahill[8],Papadimitratos[9] and Hu[6] have separately introduced detail about Wormhole

attacks in wireless networks. Initial proposals to avoid wormhole attacks propose using secure

methods of bits over the wireless channel that can be recognize only by authorized nodes. This

only defends against outside of network attackers who do not own cryptographic keys.

19

cannot be prevented by cryptographic mechanisms alone

3.1.1 Distance or Time Limiting Detection Approaches

The concept behind this approaches are intuitive, it restricts the distance or the period that

packets can traverse between nodes in network. When node in network receives packets, it will

check the transmission range of nodes or the transmission time. If a packet traverses more than

a default value, this packet is perhaps being affected by malicious attacks or goes through the

wormhole tunnel. Hu, L. and Evans et al [6] proposed a general mechanism related to this

concept called “Packet Leashes”. It add the information to the packets, this information is

designed to restrict the packet’s maximum allowed transmission distance, we called the packets are “Packet Leashes”. Two types of packet leashes were presented: Geographic Leashes and

Temporal Leashes. The first leashes, each node has to know its precise location and all nodes

have to know another node’s location information. Before sending a packet, each node adds the

information of its current location and time in the packet. When the receive node receives the

packet, it checks the packet by computing the distance to the sending node or the transmission

time of the traverse path. The receiving node can use this computing result to decide whether

the packet was transmitted through wormhole nodes. In Temporal Leashes, all nodes require

20

the packet. When receiving the packet, the receiving node compares the temporal leash of the

packet to its time, and computes the distance to the sending node by assuming the propagation

speed is equal to the light speed. As a result, it can determine if the packet traveled an overlong

distance caused by wormhole attacks. The drawbacks of Packet Leashes are that all nodes in

network need accurate time and close time synchronism; and Geographic Leashes require extra

hardware such as GPS or location service to let each node obtain its precise location.

This method is the earliest proposed method to defend against wormhole attack. The idea

of this mechanism is very simple and ordinary. However, this method requires time

synchronization or accurate location information on each node to calculate the distance

between nodes.

3.1.2 Topology detection Approaches

These methods use geometric or topology information to detect wormhole attacks. If the

analyzed results of the collected information violate the predefine situation, wormhole attacks

may occur in the network. These methods do not require time synchronization, but need more

complicated processes and message exchanges to observe and collect the information of

packets. Lazos et al. [5] proposed a topology detection approach using cryptographic

21

neighbor nodes to prevent wormhole attacks. LBK does not need any time synchronization, but

require a few additional network nodes, the guard nodes, which know their location and own

broader transmission range than the regular nodes. While establishing LBKs, all guard nodes

broadcast their fractional keys and location information to the network; and then regular nodes

collect every fractional key they received. If two regular nodes share more than a threshold

number of fractional keys, they use these keys to generate a pair-wise key. Finally, every node

generates an LBK and unicast it to the nodes which it shares with same pair-wise key. After

establishing the LBKs, each node can only communicate with their real nodes. In addition,

Lazos et al. also provide a simple mechanism, called closet guard algorithm (CGA), which

adopts the observation that a regular node should not receive fractional keys from guard nodes

that are at a distance of more than two times of the transmission range of guard nodes, to

distinguish which guards are infected by wormhole attacks.

3.1.3 Graph Theoretic and Geometric Approaches

In Geometric approaches, the nodes have to send their neighbor list to their neighbors.

The neighbors are guards to each other and monitor the transmission of their neighbors.

Like[26], LITEWWORP is geometric method proposed by Khalil et al. In the LITEWORP,

22

unreasonable actions of node are detected, the malicious counter increases. Once the

malicious counter on a particular neighbor is higher than a threshold, the neighbor revoke the

node from its neighbor list and trigger an isolation algorithm to isolate the node which is

thought as malicious. In their analysis, if the coverage of neighbors is not wide enough or too

many/less neighbors aggregate in a region, the performances both go down. The neighbors of

a node have to be kept in about 9~25 nodes, the system may work well above 90% detection

rate. However, the false alarm rates of LITEWORK are between 10% and 28% when the

neighbors number is about 17~29. As the result, we think only when the neighbors of a node

around 9 to 17, the scheme do work. And hence the nodes in LITEWORP have to monitor the

communication s of all neighbors, the energy consumption also be a problem.

3.1.4 Other Mechanisms and Protocols

Some Wormhole Attacks detection methods use the extra hardware or physical property to

detect attacks. In [6], Hu and Evans utilize directional antennas to prevent wormhole links.

Unlike our method, every node of the network is equipped with directional antennas and all

antennas should have the same orientation. Different directions called zones are sequentially

numbered and every node includes the transmitting zone at each message. A receiver hearing

23

where A, B are opposite zones. Based on information provided by neighbors that assist the

wormhole detection by acting as verifiers, every node discovers its neighbors. As pointed out

by the authors of [6], a valid verifier must exist in order for the wormhole to be detected, since

not all neighbors can act as verifiers. Finally, as noted by the authors of [6], this method can

only prevent single wormholes and does not secure the network against multiple wormhole

links [6].

3.2 Comparison of Main Mechanisms

We will discuss the main difference and conceptions of the mechanism mentioned above.

Look at Table 1, here we list the main ideas of the mechanisms, the original routing protocols

they based on or the improved, and their special assumptions, hardware or special nodes needed

in the mechanisms. Finally, we compare their computational and control overheads.

First, Distance or Time Limiting Detection Approaches use the packets transmission

distance to judge a node. This way can detect the Wormhole attacks effectively due to

transmission distance cannot forge except that malicious nodes have capacity of moving the

node’s location rapidly. However, this solution needs accurate location information or precise

clock synchronization. And this way assumes the nodes are not movable, this drawback greatly

24

pairs have to be encrypted by local broadcast key of the sending end and decrypted at the

receiving end. As a result the time delay in each node is extended a lot. Finally, the Statistical

analysis methods use a central authority (CA) or coordinate node to monitor the Ad-hoc

networks. If the central node can collect and analysis the network information, it will aware the

behavior of Wormhole attacks as soon. But, practically, most ad-hoc network routing protocols

are tend to non-central authority. So, this statistical solution requires particular routing

protocols or specific hardware to implement.

Protocol

Description

Drawback

Hu and al. 2003 Use of packet leaches

With geographical and temporal information

requires synchronized clocks and GPS equipped devices

L. Hu and al. 2004 Use the direction of the antenna Of the neighbors

use of directional antenna

R. Maheshwari and al. 2007

Search for forbidden structure caused by the wormhole

Difficulty to compute a parameter to determine forbidden structure

25

Chapter 4 The Proposed Protocol:

AR-AODV Routing Protocol

AR-AODV protocol presents Auto Resilient AODV routing protocol. In this Chapter, we

introduced the AR-AODV protocol include design conception and the future. In Section 4.1, we

first describe our observation about the Wormhole Attacks using AODV routing protocol; and

later we proposed our design protocol against Wormhole Attacks. Also, we described

particularity how to implement our AR-AODV routing Protocol in final.

4.1 Observation of Wormhole Attacks on AODV

We proposed a routing protocol for preventing the Wormhole Attack on Ad-hoc Network,

called AR-AODV (Auto resilient Ad hoc On-Demand Distance Vector Routing). This Protocol

is based on AODV (Ad hoc On-Demand Distance Vector Routing) protocol. The original

AODV routing protocol cannot prevent from wormhole attack. Our Protocol, AR-AODV, uses

the node hop numbers difference to detect the Wormhole Attack.

26

discuss our observation about the characteristic of the network it was be infected by Wormhole

Attack.

a. How Wormhole Attack take place on AODV

The major mission for malicious node intended to execute Wormhole Attack is how to

get many chances to achieve the maximum packages on network. As soon as malicious node

can obtain more and more packages, the Wormhole attack’s destruction will be more serious.

On Ad-hoc network, routing cost is an important concern. So if any router can decrease the

routing cost, it will get many chances to achieve the packages. For this reason, Wormhole

attacks will try all the means to decrease the routing cost. Next, we will introduce how

Wormhole attacks decrease the routing cost on AODV.

Protocol such as AODV discover and decide routing path by comparing the hop counts in

the RREQ package. Hop Count is the number of hops from the Originator IP Address to the

node handling the request. A typical Wormhole attack occurred on AODV is possible by Type D G Reserved Hop count

Broadcast ID

Destination IP address

Destination sequence number

Originator IP address

27

modification of the hop count field in route discovery messages RREQ. When routing decisions

cannot be made by other metrics, AODV uses the hop count field to determine a shortest path.

In AODV, malicious nodes can increase the chances they are included on a newly created route

by resetting the hop count field of the RREQ to zero.

28

Figure 4-2 wormhole attack on AODV

b. our observation about the characteristic of Wormhole attack

Based on the above described, we observation the changes in packets. There is a

difference about the hop count between before packets through the wormhole node and after

packets through the wormhole node. According to the clues, we monitor the hop count of

packets immediately. Therefore, when one hop counts of RREQ with the dramatic changes, we

can take appropriate measures to prevent attack in time. However, it is not enough to guard

against such wormhole attacks. The measure of monitoring hop count can only determine

29

wormhole node. Our goal is not only discover attacks but also auto resilient the network

barriers, so we must do more to meet our requirements. Reviewing the packets message of

AODV, we think there are not enough clues to identify specific wormhole node. For this reason,

we have added one column to the RREQ packet. This column is recorded the path history from

source node to destination node. After adding this path column to track the path and collecting

the monitor of hop counts, we can determine whether Wormhole Attacks has occurred, and

more importantly, this method can identify the wormhole node specifically.

4.2 The Proposed Protocol: AR-AODV

In this section, we will bring up our ideas to solve the security problem as mentioned above.

In our proposed protocols, we can detect the malicious node altering the RREQ’s hop counts.

30

Figure 4-4 Reverse RREQ

4.2.1 Design Conception

First, our AR-AODV protocol will add one column into the RREQ package. The original

RREQ packages on AODV protocol do not contain the any information about the routing path

history. The fewer packet size perhaps can increase routing efficiency and speed up the duration

of data transmission. However, we find if the packages do not contain the routing information,

it will become a huge security loophole on Ad-hoc network. For this reason, we slightly modify

31

Type D G Reserved Hop count

Broadcast ID

Destination IP address

Destination sequence number

Originator IP address

Originator sequence number

Routing history

Table 2 Traceable RREQ

Second, our AR-AODC protocol will add a new routing packet type: Reverse RREQ.

Comparing with the RREQ packets sent from source node, Reverse RREQ packets are sent

from destination node, and traced back along the RREQ’s path to find the malicious node.

Reverse RREQ are like RREQ, when destination nodes send reverse RREQ package, the node

should assign a goal. After assigning the goal, Reverse RREQ package will go along the old

path to count the routing cost. And we find, we can detect malicious node if we send

continuously two reverse RREQ packets that target nodes are different and malicious node is

32

4.2.2 Route Establishment

Figure 4-5 Illustration of route establishment

1. Send Request

When one node needs to send a message to another node that is not its Neighbor it broadcasts a

Route Request (RREQ) message. The RREQ message contains several key bits of information:

the source, the destination, the lifespan of the message and a Sequence Number which serves as

a unique ID.

2. Receive request

When Node 1’s Neighbors receive the RREQ message they have two choices; if they know a route to the destination or if they are the destination they can send a Route Reply (RREP)

33

Neighbors. The message keeps getting rebroadcast until its lifespan is up.

3. Forward

When one node receives a RREQ message and it does not know how to route to the destination,

it will forward this RREQ message to neighbors.

4. Send Reverse Request

When destination node receive the RREQ message, it should check the RREQ whether the

routing path contain Wormhole nodes.

5. Send Reply

If setp.4 has checked and the routing path does not contain wormhole nodes, then destination

node will send out a RREP message to source node from the request routing path.

6. Receive Reply

When source node receives the RREP message, in our methods, it means that the path is safe.

4.2.3 Route Maintenance

When wormhole nodes have been detected, the next most important thing is how to disable

the wormhole nodes. Actually, it is not as simple as wired network which can easily find the

physical location of any nodes or just disable the attack node from the router. In Wireless

Ad-Hoc network, node can easily change its location and we cannot disable one node from a

34

Yes

No

node in Ad-Hoc network directly, we adopt an indirect approach to achieve the purpose of the

prohibition of attack node. Reviewing the AODV routing protocol, it define a HELLO message

to maintain the network. We used this HELLO message to disable the Wormhole attack node.

HELLO message can be communicate with directly are considered to be neighbors. In our

methods, we add one column to record the list of wormhole attack nodes. When nodes in the

network receive the HELLO message, it will check the list of wormhole attack nodes and add

the list to the routing table. Our mechanism is when a node receives RREQ from neighbors, it

should query the routing table to judge whether the neighbor can be trusted. If the neighbor’s id

is in the list of wormhole attack nodes, the AR-AODV will automatically add a weighted value

to the hop count number of RREQ message. The effect of doing so is to reduce the chance of the

routing path include wormhole node be selected.

In the route table

Broadcast RREQ packet from source

Intermediate nodes receive the RREQ

Generate a REPLY

Cache Src. Address, Dst. Address

35

Yes

Send REPLY message

Hop count +1

Destination node

Send Reverse RREP The existence of wormhole nodes

Yes

No

Drop the RREQ package

Route maintenance Send Hello message

36

Chapter 5 Simulation and Results

In this Chapter, we will evaluation our AR-AODV protocol through simulations. The first

section will explain out simulation environment, include the simulation tool ns-2[7]. The

second section, we will show three scenarios of simulations to evaluation the AR-AODV

including its performance, the rate of detecting Wormhole Attacks. Finally we will show the

simulation results.

5.1 Simulation Environment

In our simulation, we use the NS-2 as the simulation tool. NS -2 is a discrete event

simulator targeted at networking research. Ns-2 provides substantial support for simulation of

TCP, routing, and multicast protocols over wired and wireless (local and satellite) networks.

Also, we chose NAM as our network animation tool. Nam is a TCL / TK based animation tool

for viewing network simulation traces and real world packet traces. It is mainly intended as a

companion animator to the ns simulator.

37

transmission range of simulator are a fixed value of 0.25m. In order to simulate the more

realistic situation, we evaluate two types of Ad-hoc network model. One is smaller network

used usually in room or classroom; the other one is larger network used in indoor buildings. The

smaller network we define the length and width of 10 meters with 50 nodes; the larger network

model we define the length and width are 50 meters with 1000 nodes.

CPU Intel(R) Core(TM2) Duo CPU T7250 2.00.GH

Memory 4 GB

OS Microsoft Windows XP

NIC Inter(R) PRO/1000 Ethernet

Hard Disk SCSI 250G

38

Simulation parameter value

Radio-propagation model Propagation/TwoRayGround

MAC type 802.11

Routing protocol AODV/AR-AODV

Time of simulation end 300.0s

Antenna model OmniAntenna

Table 4 Simulation parameter

39

Figure 5-2 large grid (50M*50M, 1000 nodes)

5.2 Simulation and Results

In this section, we show the simulation results of AR-AODV protocol and the analysis of

simulation outcome. In our simulation scenarios, we focus on two parts: detection rate and

overhead. We will use charts to show it.

5.2.1 The impact of wormhole attack

The main purpose of this experiment is to measure the detection rate of routing protocols.

We define a measurement criterion: Wormhole attack impact rate. It means the probability of

data packets through the wormhole attack node. The higher of this impact rate of routing

40

three cases. The first case that T is 0 means the original AODV routing protocol. This case will

have a high impact rate due to AODV routing protocol does not have defense mechanism to

wormhole attack. In other two cases, they all use the AR-AODV routing protocol. But these two

cases are not the same as the timing of the use of detecting methods. When T is”average path

length”, it will check the RREQ packets just when hop count number less than average path

length. However, when T is ”the longest path length”, it will check the RREQ packets all the

time.

41

Figure 5-4 impact rate (large grid)

5.2.2 Overhead of AR-AODV

The main purpose of this experiment is to assess the overhead of AR-AODV. In 5.2.1

section, the simulation results show that AR-AODV can suppress the wormhole attacks.

However, we evaluate a routing protocol also should consider the overhead of the methods. If

the overhead of new approach are much heaver than original protocol, it is not a

high-perfomance method.

In first part, we show the overhead about the latency of RREQ packets. The latency means

the speed of finding a routing path form source node to destination node. If the value of latency

42

the original routing protocol AODV has the smallest delay. The latency of AR-AODV routing

protocol are more than twice as high as AODV. However, the latency will gradually become

smaller with the time. This is due to that AR-AODV should send Reverse RREQ packets at first

time if the routing table has no cache data, but next time the AR-AODV can query the routing

table data, it will reduce the delay time. The latency will reduce to just higher than the latency of

AODV a little. Extra latency is because AR-AODV should do some routing maintaining

actions.

43

Figure 5-6 latency of RREQ packets (large grid)

In second part, we show the number of RREQ packets in network. When AODV routing

protocol is in the initial stage, it should send the RREQ packets to discover network topology.

Our AR-AODV routing protocol also uses the initial stage to detect the wormhole attack. The

purpose of this experiment is to compare the overhead of number of RREQ packets. The lower

number of RREQ packets will be able to mitigate the load of network. From the simulation

results, the original routing protocol AODV has the smaller number of RREQ packets in the

beginning, it is three times less than AR-AODV. However, the number of RREQ packets will

44

the network after the network initial stage. If the routing table has cached the routing data, the

RREQ packets will never be sent out. Another reason about the our protocol has smallest

number of packets at final is because some nodes have be disabled, it also will reduce the

number of RREQ packets.

45

46

Chapter 6 Conclusion and Future Work

6.1 Conclusions

In this thesis, we studied the effect of the wormhole attack in mobile Ad-Hoc networks

(MANETS) and analyze the security issues of MANETs. We proposed a new on-demand

routing protocol, Auto-Resilient Ad hoc On-Demand Distance Vector Routing (AR-AODV),

to against wormhole attacks on mobile Ad-Hoc networks (MANETs) and recover the network.

AR-AODV utilizes Reverse RREQ packets to detect wormhole attacks and disable the

wormhole node using Hello packets. The original AODV routing protocol has no mechanisms

to against wormhole attack, the malicious can tamper the hop number of RREQ to cause huge

amounts of packets gathering into wormhole node. In our proposed method, it can effectively

prevent the wormhole attack on MANETs. Also, the Hello messages can auto disable the

wormhole node, we call this is “Auto-resilient”. Indeed, the simulation results showed that

wormhole nodes are disappeared gradually.

6.2 Future Work

Our proposed routing protocol is mainly used in AODV routing protocol. However, we

47

step of finding the routing path and routing maintain. In the future, our method can be ported

to other routing protocols like: DSDV (Destination-Sequenced Distance-Vector Routing), this

routing protocol used Bellman–Ford algorithm to find the routing path; DSR (Dynamic

Source Routing), this routing protocol also has the Routing Request packets, so we can port

48

Reference

[1] (AODV) Routing,” IETF MANET Working Group, Internet Draft 2003. Computer, vol.

35, no.10, pp. 54-62, 2002.

[2] Lionel M. Ni , Philip K. McKinley, A Survey of Wormhole Routing Techniques in Direct

Networks, Computer, February 1993 , v.26 n.2, p.62-76

[3] Y. Hu, A. Perrig, D. Johnson, Packet leashes: a defense against wormhole attacks in

wireless networks, in: IEEE Annual Conference on Computer Communications

(INFOCOM), 2003, pp. 1976–1986.

[4] W. Wang, B. Bhargava, Visualization of wormholes in sensor networks, in: Proceedings

of the ACM workshop on Wireless security (Wise’04), 2004, pp. 51–60.

[5] L. Lazos, R. Poovendran, C. Meadows, P. Syverson and L.W. Chang, Preventing

wormhole attacks on wireless ad hoc networks: a graph theoretic approach, IEEE

Wireless Communications and Networking Conference (WCNC05) 2 (13–17) (2005), pp.

1193–1199

[6] Hu, L. and Evans, D. 2003b. Using directional antennas to prevent wormhole attacks. In

Proceedings of the 11th Network and Distributed System Security Symposium.

131--141.

49

[8] B. Dahill, B. N. Levine, E. Royer, and C. Shields, “A secure routing protocol for ad-hoc

networks,” Electrical Engineering and Computer Science, University of Michigan,Tech. Rep. UM-CS-2001-037, August 2001.

[9] P. Papadimitratos and Z. Haas, “Secure routing for mobile ad hoc networks,” in SCS

Communication Networks and Distributed Systems Modeling and Simulation

Conference (CNDS), 2002.

[10] Y. C. Hu, A. Perrig, and D.B. Johnson, “Packet leashes: a defense against wormhole

attacks in wireless networks,” in Proceedings of the 22nd INFOCOM, pp. 1976-1986, 2003.

[11] Radha Poovendran , Loukas Lazos, A graph theoretic framework for preventing the

wormhole attack in wireless ad hoc networks, Wireless Networks, v.13 n.1, p.27-59,

January 2007 [doi>10.1007/s11276-006-3723-x]

[12] Khalil, I., Bagchi, S., and Shroff, N. B. 2006. MOBIWORP: Mitigation of the wormhole

attack in mobile multihop wireless networks. In IEEE International Conference on

Security and Privacy in Communication Networks.

[13] Xiaoyan Hong, Kaixin Xu, and Mario Gerla. Scalable routing protocols for mobile ad

hoc networks. 2002.

[14] Elizabeth M. Royer and Chai-Keong Toh. A review of current routing protocols for ad

50

Institute of Technology, USA, 1999.

[15] C. Santivanez and R. Ramanathan, Hazy Sighted Link State (HSLS) Routing: A Scalable

Link State Algorithm Cambridge, MA: BBN Technologies, Aug. 2001, BBN Tech.

Memo BBN-TM-1301.

[16] Zygmunt J. Haas , Marc R. Pearlman, ZRP: a hybrid framework for routing in Ad Hoc

networks, Ad hoc networking, Addison-Wesley Longman Publishing Co., Inc., Boston,