OPNET_MODELER_MANET Attacks Thesis

Mobile Ad hoc Network Security Issues By Sheraz Salim Student ID 20388626

Presented to Faculty of School of Computing, Engineering and Physical Sciences University of Central Lancashire

In Partial Fulfilment Of the Requirements For MSc Computer Networking

Name: Sheraz Salim

Page 1

Master Project CO4804

Acknowledgement I always thank to ALL-MIGHTY ALLAH, who gave me power and courage to complete this project. I would like to thank Zaki-El-Haroun for his supervision during this project and Helen Campbell Campbell for supporting supporting and gaudiness during this tough time. I also have good regards for the facility of school of Computing Engineering and Physica Physicall Science Science,, Univers University ity of Central Central Lancash Lancashire ire includin including g Chris, Chris, who always always suppo support rt us for the the simula simulati tion on softw software are probl problem ems. s. And also also thank thankss to OPNE OPNET T technologies software. I would also like to express my sincere thank to my friends who support me all this time during thesis first and foremost, Noman Qazi and Hamid Khan. Finally, Finally, I would like to dedicate dedicate this thesis to my parents, brother and sister. Without Without their support it wasn’t possible to complete this project.

Name: Sheraz Salim

Page 2


Abstract Wireless network are getting common around over living atmosphere and user are enjoying the wireless connectivity irrespective of their geographic position. Mobile ad hoc network also know as MANET is a fast growing technology which give connectivity where existing technologies technologies communication failed due to catastrophic catastrophic failure. failure. MANET MANET technology technology received a hug hugee atte attenti ntion on from from the resea researc rcher herss to make make them them bett better, er, reli reliabl ablee and and valua valuable ble in communication environment, MANET nodes are highly self organized with a collection of few mobile node they formed formed a tempora temporary ry network network.. All these fantasiz fantasizing ing features features,, and marvellous connectivity with other mobile node are making a enormous network, despite when it come to security it pose a large number of threats to the network. In order to provide secure transmission, researcher proposed many security measures within the network i.e. secure routing protocol and cryptography. The scope of this project is to study the effects of Pulse Jammer attack and Misbehavior nodes using Proactive routing protocol i.e. Optimized Link State Routing Protocol (OLSR), Reactive routing protocol Ad Hoc On Demand Distance Vector (AODV) and Geographical Routing Protocol (GRP). The impact of attack on MANET performance performance is evaluated evaluated finding out which protocol is more vulnerable to these attacks. The measurements were taken in light of throughput, network load, delay and packet delivery ratio. ratio. The simulation simulation work is done in Optimized Network Engineering Tool (OPNET).

Name: Sheraz Salim

Page 3


Table of Contents Con tents Table of Contents................................ .................................................................... ................................................... ...............4 4 CHAPTER NO 1.................................... ....................................................................................... ...................................................7 7 Mobile Ad hoc Wireless Network....... .............. .............. ............... ............... .............. .............. .............. .......... ...7 7 Introduction...................................................... ............................................................................... .................................... .............7 Aim of thesis................................... ....................................................................... ................................................... ................. ..10 1.3

Project Structure.................................. ........................................................................ ......................................10 10

CHAPTER NO 2.................................... ..................................................................................... .................................................10 10 2.1 Physical Layer Attacks............................... ................................................................... ....................................10 10 Jammer Attack............................... ................................................................... ............................................... .................. .......11 11 Interference Problem.................................... ........................................................................... .......................................12 12 Eavesdropping Eavesdroppin g Attack................................... .......................................................................... .......................................12 2.2 Data Link Layer Attacks.................................... .............................................................. ............................. ...12 IEEE 802.11 attack.................................. ..................................................... ............................. ..................... ............... ....13 Virtual Jamming attack 802.11.................................. ............................................... ........................ ............. ..13 13 2.2.3 Traffic Monitoring and Analysis Attack ..................... ............................... .................. ........14 14 Single adversary attack (SAA)............................... ................................................ ............................ ............. ..14 14 Colluding adversaries attack (CAA)......................................... .................................................... .............14 2.3 Network Layer Attack...................................................... ................................................................ .............. ....15 15 2.3.1 Routing Protocol......................................................................... Protocol.........................................................................15 15 Proactive (Table Driven Routing Protocols)........................ ................................... ................. ......16 16 Reactive (On- Demand Routing Protocols) .........................................16 .........................................16 Other Routing Protocol........................................................................ Protocol........................................................................17 17 Comparison of Proactive, Reactive and Hybrid Routing Protocol in Comparison MAN ANE ET 17 Attack on Network Layer ............................... ................................................... ............................... .................. .......18 18 Routing attacks................................ .................................................................... .................................................... ................19 19 Routing table overflow attack....... .............. .............. .............. ............... ............... .............. .............. ........... ......19 Routing table poisoning attack.................................... ........................................................ ........................ ....19 19 Route cache poisoning attack........................................ ................................................... ..................... ..........19 19 Packet forwarding attacks or Advance attacks....... ............... ............... .............. ........... ....... ...20 20 Black hole attack............................... .......................................................................... .................................................. .......20 20 Wormhole Wormhol e Attack............................... ............................................................................... .................................................. ..20 Jellyfish attack.................................. ...................................................................... .............................................. ............... .....23 23 Gray holes attacks.................................... ............................................................................... ...........................................23 23 Name: Sheraz Salim

Page 4


Byzantine attack.................................... ........................................................................ .............................................. ..........24 24 Packet Replication.................................... ..................................................................... ........................................... ..........24 24 Rushing attack.................................... ........................................................................ ................................................. .............24 24 2.4 Transport layer Attack.................................. ...................................................... ............................... ............. ..24 Replaying acknowled acknowledgement gement....... .............. .............. ............... ............... .............. .............. .............. ............ .......24 SYN flooding attack .................................. ....................................................... ................................ ..................... ............25 2.4.3 Session hijacking................................ .............................................................. ........................................ ..........25 25 2.5 Application layer Attacks................................... ................................................ ........................ ............... ....25 25 2.5.1 Applicatio Application n layer attacks MANET.............................................. ................................................ ..26 26 Malicious attack ................................................................................ .................................................................................. ..26 Repudiation attack............................... ........................................................ .................................... ..................... ............26 2.6 Cross layer security Attack.................................. ............................................... ........................ ............. ..27 27 Passive attack................................. ..................................................................... ..................................................... .................27 27 Chapter NO.3................................. ..................................................................................... ...................................................... ..29 29 3.1 MANET Security Services................................... ...................................................... ............................ .........29 29 Authentication Authenticatio n......................................................................... .................................................................................... .............29 Confidentiality Confidential ity ................................................ ............................................................... ......................... ..................... .............30 Integrity 30 Availability Availabili ty.................................. ...................................................................... ................................................. ..................... ........30 30 Non-Repudiation Non-Repudiati on................................ .................................................................................. ..................................................30 30 CHAPTER NO.4.................................... ..................................................................................... .................................................31 31 IMPLEMENTATION IMPLEMENTAT ION ............................................... .......................................................... ...................... ..................... ............31 4.1 Discrete Event Simulator (DES)..................................................... .....................................................31 31 4.2 Continuous simulatio simulation n................................ ............................................................. ................................... ......31 31 4.3 Test bed Labs............................... ................................................................... .............................................. ............. ...31 31 4.5 Simulation tool............................................................................... tool...............................................................................32 32 4.6 The Architecture of OPNET............................... ......................................... .................... ................... .........32 32 4.7 Intelligent Pulse Jammer Attack Model.......................................... Model..........................................33 33 4.8 Misbehavior Node Attack Model.................................................... Model....................................................33 33 4.9 MANET Network Model.................................................................. Model..................................................................34 34 4.10 Applicatio Application n Configuration Setting ................................................ ................................................34 34 4.11 Profile Configuratio Configuration n Setting....... .............. .............. .............. ............... ............... .............. ........... ...... ..35 35 4.12 Mobility Configuration Setting............................................ ..................................................... .........36 36 4.13 Node Model Traffic Packet setting............................................ ............................................... ...36 36

Name: Sheraz Salim

Page 5


4.14 Creating Intelligent Pulse Jammer Node Model........................... Model...........................38 38 4.15 Creating Misbehavior Node Model............................................... Model...............................................39 39 4.16 Scenario 1.................................. ...................................................................... .................................................. ..............39 39 4.17 Scenario 2.................................. ...................................................................... .................................................. ..............40 40 4.18 Scenario 3.................................. ...................................................................... .................................................. ..............40 40 4.19 Scenario 4.................................. ...................................................................... .................................................. ..............41 41 4.20 Scenario 5.................................. ...................................................................... .................................................. ..............41 41 4.21 Scenario 6.................................. ...................................................................... .................................................. ..............42 42 4.22 Performance Metric and Analysis............................. ........................................ ................... ........42 42 CHAPTER NO.5.................................... ..................................................................................... .................................................44 44 5.1 RESULT AND ANALYSIS.................................. .................................................................. ................................44 44 5.2 SCENARIO 1.................................... ................................................................................... ...............................................44 44 5.2.1 Throughput.................................... ................................................................................. .............................................45 45 5.2.2 Network Load.................................. .................................................. ........................... ..................... ................ ......46 46 5.2.3 Delay..................................... .............................................................................. ................................................... ............ ..47 47 5.2.4 Packet Dropped................................. .................................................................. ......................................... ........47 47 5.2.5 OLSR routing Hello packet.................................. .......................................................... ........................48 48 5.3 Scenario 2................................ .................................................................... ...................................................... ..................49 49 5.3.1 Throughput.................................... ................................................................................. .............................................50 50 5.3.2 Network Load.................................. .................................................. ........................... ..................... ................ ......51 51 5.3.3 Delay .........................................................................................52 5.3.4 Packet Delivery Ratio............................................... .......................................................... .................. .......52 52 5.4 Scenario 3................................ .................................................................... ...................................................... ..................53 53 5.4.1 Throughput.................................... ................................................................................. .............................................55 55 5.4.2 Network Load.................................. .................................................. ........................... ..................... ................ ......55 55 5.4.3 Delay..................................... .............................................................................. ................................................... ............ ..56 56 5.4.4 Retransmi Retransmission ssion Attempts........................................ ................................................... ................... ........57 57 5.4.5 Route Discovery Time....... .............. .............. .............. ............... ............... .............. .............. ............. ........ ..57 57 5.5 Scenario 4................................ .................................................................... ...................................................... ..................58 58 5.5.1 Throughput.................................... ................................................................................. .............................................59 59 5.5.2 Delay..................................... .............................................................................. ................................................... ............ ..60 60 5.5.3 Packet delivery ratio (PDR).................................... ......................................................... .....................61 61 5.5.4 Network Load.................................. .................................................. ........................... ..................... ................ ......62 62 5.6 Scenario 5................................ .................................................................... ...................................................... ..................63 63

Name: Sheraz Salim

Page 6


5.6.1 Throughput.................................... ................................................................................. .............................................64 64 5.6.2 Delay..................................... .............................................................................. ................................................... ............ ..65 65 5.6.3 Network Load.................................. .................................................. ........................... ..................... ................ ......66 66 5.6.4 Packet Delivery Ratio............................................... .......................................................... .................. .......66 66 5.7 Scenario 6................................ .................................................................... ...................................................... ..................67 67 5.7.1 Throughput.................................... ................................................................................. .............................................68 68 5.7.2 Network Load.................................. .................................................. ........................... ..................... ................ ......69 69 5.7.3 Packet Delivery Ratio............................................... .......................................................... .................. .......70 70 5.8 Analysis................................ ........................................................................... ...................................................... .............. ...71 CHAPTER NO.6.................................... ..................................................................................... .................................................75 75 6.1 Conclusion................................... .............................................................................. .................................................. .......75 75 6.2 Findings................................. ..................................................................................... ........................................................ ....76 6.3 Future Work................................... ................................................................................... ................................................77 7. References................................... ....................................................................... .................................................... ................78

CHAPTER NO 1 Mobile Ad hoc Wireless Network Introduction The advent of wireless technology is one of the biggest breakthroughs of modern technology. This technology enables the users to utilize the freedom of movement and the use of the equipment while on the move. Initially the use of wireless technology was restricted for the military purposes to provide cutting edge to the defence forces. However, during the 1980s Name: Sheraz Salim

Page 7


there was an explosive growth of this technology. For general public the first interaction with the wireless technology was the mobile phone. Initially the technology was expensive to use and difficult to maintain. But with the rapid advancement advancement in this field the technology became affordable and with the implementation of fourth generation of wireless networks the entire concept has changed. In recent year Mobile Ad hoc Network (MANET) (MANET) seek much attention from industry and academic researchers from all around the globe. This technology has come with its own flavour as they are easy to deploy in disaster areas and for emergency operations due to its nature of design and operation, such as an environment where a collection of few mobile nodes can form a temporary network without a need of any centralized networks, with the collaboration of each other nodes they sent and forwarding packets to destination nodes. Each node act host and router at the same time to forwards the packet. Mobile Ad hoc Network do not rely on wired network routing protocols. Mobile Ad hoc Network uses Ad hoc routing protocols, these protocols are entirely different than tradition wired network as routing path breaks and connected due to self organized nodes to discover multi hop routes through network to any other node. This functionality is also known as infrastructure less. Wireless network provide connectivity between end nodes i.e. PDA, Notebooks, Mobile phones etc, without the use of wires. The transmission occurs on electromagnetic waves, know as radio waves. Wireless network can be combination of switches, routers and hosts without any wired connected to each other, they must functional wirelessly. A router is responsible for packet forwarding in the network and host are source or destination of the data flow. Wireless network giving edge compare to traditional networks as there is no physical physical cable required for data transfer. Wireless communication communication gives a freedom to host to move freely in wireless communication zone network. To accompany this freedom wireless communication relay on channel know as radio frequency RF, Wireless nodes are attached with a centralized fixed network such as Access Point (AP). The Access Point (AP) acts as a middle man between the wireless wireless nodes, as wireless nodes are on the move from one place to another AP keeps them connected to a wireless network. The existence of these networks is not feasible as fixed network also know as infrastructure is limited support to the adaptability of wireless network. Therefore this technology is not reliable where no fixed infrastructure is present like rural areas. In recent advancement technologies such as Bluetooth and IEEE 802.11 launched a paradigm of wireless system called as Mobile Ad hoc Networks (MANET). The speciality of this technology is that it could be managed even in lack of fixed infrastructure. This technology is efficient, effective, quick, and easy to deploy. MANET consists of independent mobile nodes

Name: Sheraz Salim

Page 8


connected by wireless medium, each mobile node acts host and operates as an end system, and also acts as router for all nodes in the network. MANET are heterogeneous as path between each single pair may have multiple links, each mobile node is free to roam while communicate with other nodes and self organize. This forms a relationship of different kind of links to become a part of the same network [1]

Figure 1.1 Mobile Ad hoc Network.[32] While Mobile Ad hoc network have many advantages advantages over the traditional traditional wired network, but when it come to security it pose a immense set of threats. Firstly Mobile Ad hoc network face the threat of secure communication in wireless medium. The MANET connectivity poses a great threat to radio frequency in fallowing ways; [2] 

The wireless wireless medium medium has no complet completee or no certain certainty ty visible visible limita limitation tionss

outside of which stations are known to be incapable to obtain network frames. 

The wireless medium is defenceless from outside the channel signals.



The wireless medium is considerably less undependable then wired media



Thee phen Th phenom omen enaa of hidd hidden en term termin inal al and and expo expose sed d term termin inal al may may occu occur r

consistency.

Secondly, nodes having no security on the network may face threats from active attack as impersonation to passive attacks as eavesdropping which lead network to deceased and degrade the performance and resources. Vulnerable malicious node attempts to intercept the Name: Sheraz Salim

Page 9


radio wireless transmission and modify the packets without the knowledge of authorized nodes. Likewise, various attacks Denial of Service could be launch easily to start sending packets through vulnerable nodes which give false update information to disturbed and poison the routing information in the network. As result the network is destroyed and the motive of attack will be successful.

Aim of thesis It is try to provide as much as information in regards of malicious node, the techniques, methodology methodology and tools which are assume as attacker approach to harm or hijack the network resources in light of few attacks that are analysed on OPNET Modeler 14.5, We believe implementing few number of attacks on MANET with having enough knowledge of the attack as prospective of attacker thinking, can help to investigate the weakness of the network and have better understanding to secure the network.

1.3

Project Structure

The project is divided in to seven chapters. Chapter 1 give brief description of MANET technology and the problem statement in light of security issues. Chapter 2 presents the theoretical concept of security vulnerabilities of each layer in MANET i.e. Physical layer, Data link layer, Network layer, Transport layer, Application layer and cross layer security atta attacks cks.. Chapt Chapter er 3 presen presents ts the the MANE MANET T securi security ty servi service ce in term termss of authen authenti ticat cation ion,, confidentiality, confidentiality, integrity, integrity, availability availability and non-repudiation. non-repudiation. Chapter 4 explains explains the simulation tool OPNET Modeler 14.5 along with experimental design, implementation of the network we created to analyse in terms of security; application configuration, profile configuration, and mobility configuration are defined. Pulse jammer node model and Misbehavior nodes model is created to implement them on network. Different routing protocols are carried out in each scenario to generate the traffic. This chapter mainly focus on the simulation and nodes settings. Chapter Chapter 5 is about the performance performance metrics throughput, delay, network load, packet delivery ratio and retransmission attempts results have been discussed. Finally chapter 6 presents the conclusion, finding and future work.

CHAPTER NO 2 2.1 Physical Layer Attacks Physical layer is the lowest layer of OSI reference model. This is commonly shortened as “PHY”. “PHY”. The physical layer has a priority compare to other layer of OSI/ISO reference model

Name: Sheraz Salim

Page 10


as it is the only layer where data is physically moved across the network. The physical layer function is encoding, signalling, data transmission, reception and physical network design. On wireless networks it functions as radio transceiver, and packet transmission through physical medium. The function of physical layer on wireless is to transmit the single in the radio wireless waves to its destination or to finding path with collaboration of other nodes which are present in the network. When physical layer receive the single it pass it to other lower layer. Radio wireless signal in MANET are highly vulnerable on physical layer, as wireless wireless radio signals travel throughout throughout the open air as having their limited frequency range. Consequently Consequently radio wireless communication communication is easy to jam or intercept intercept as its nature of using open airwaves. Any malicious node can hear the wireless signal and disrupt the transmission transmission in the network. As many attacks are executed and widely discussed in literature on wireless network, thus physical physical layer is one of important important layers of OSI reference model to secure in order to protect the radio signal from many malicious malicious attacks that can launch attack on this layer. Launching Launching attacks on physical layers are Jamming transmission, interference, eavesdropping is mostly applied and in common practice on this layer. Attacker goal is to weak the strength of radio signal but also jamming the transmission to bring the network resources down.

Jammer Attack Jamme Jammerr atta attack ck is commo commonly nly used used to wipe wipe out the transm transmis issio sion n on the the targe targett wire wireles lesss networks. Jammer attack responsibility is to stop nodes from sending and receiving packets on network. Jammer function on physical layer to generate the packet at a high constant rate on wireless medium to keep the medium busy, consequently node on network can’t access the wireless medium due to the noise jammer created in wireless medium. In this attack, the malicious node hire a device called Jammer which is use to observing the wireless radio frequency, frequency, until attacker match the frequency at receiving node receiving the signal from the sender in the network. Attacker matches the frequency with the help of Jammer frequency device of the targeted networks and transmits radio signals with generating a continuous high radio frequency (RF) it is powerful signal that overwhelmed within the range of network transmission. Subsequently messages are lost due to high noise in the spectrum. Jammer attacks are considered as brutal force attack. Jammer attacks can be classified as constant jammer, deceptive jammer and random jammer.

Name: Sheraz Salim

Page 11


Interference Problem In wireless network interference has seen commonly, it occurs when set of nodes clash with other devices which are using the same Radio Frequency (RF). Each MANET node station use 802.11 network card to transmits transmits packet in the network when no other node transmitting the packets. If any other node happens to sending packet, other node will have to wait to get the wireless medium free. Radio Frequency (RF) occupies the unwanted interference Radio Frequency signals that interrupt the normal operations as 802.11 protocol defined by IEEE get interfering Radio Frequency single of sufficient amplitude and frequency which can appear as bogus 802.11 nodes transmitting a packet. This cause genuine 802.11 station to wait for indefinite periods of time until interfering signal goes away. The sources of Radio Frequency Interface which may cause the problem are 2.4 GHZ this could be any interference single including microwaves, wireless phones, and Bluetooth devices etc.

Eavesdropping Eavesdropping Attack Eavesdropping attack is the process of gathering information by snooping on transmitted data on legi legiti timat matee netw network ork.. Eaves Eavesdro drop p secret secretly ly overhe overhear ar the trans transmi missi ssion. on. Howe However ver,, the the informa information tion remain remainss intact intact but privacy privacy is comprom compromised ised.. This This attack attack is much much easier easier for malicious node to carry on as evaluate to wired network. Eavesdropping attack in MANET shared the wireless medium, as wireless medium make it more vulnerable for MANET malicious nodes can intercept the shared wireless medium by using promiscuous mode which allow a network device to intercept and read each network packet that arrives. The attacker node intercepts the transmission as every MANET node is equipped with transceiver in range of the communication which can be decode by means of malicious node to target the authorized node on the network, malicious node can obtain the sensitive information etc, modify the routing route or poison the routing table. This can seriously harm the network resource and degrade the network performance. There are number of hardware tools that allow promiscuous mode, such as Prisme2 network card with using eavesdropping and high power antenna to intercepting wireless traffic.

2.2 Data Link Layer Attacks Data link layer is second lowest layer of OSI reference model; model; it is commonly known as link layer. It ensures the reliable communication link between neighbour nodes, data link layer define the different network and protocol characteristics, i.e. logical link control (LLC), Media Access Control (MAC), Data Framing, Addressing, and Error Detection and Handling. In wireless networks MANET nodes provide multipoint peer-to-peer network architecture.

Name: Sheraz Salim

Page 12


Their connectively is maintain by one hop among to the entire network by using link layer protocol and the network layer protocol extent the connectivity to other nodes in the network. Malicious nodes can launch attack on link layer by distracting the collaboration of protocol of this layer. [3]

IEEE 802.11 attack The vulnerability on DoS attacks in link layer IEEE 802.11 has been investigated malicious node has to keep the most of binary exponential back off scheme. In order to happen as expected malicious node damage frame easily by putting few extra bits on the frame and overlook the ongoing communication. The mechanism of this procedure is understood when IEEE 802.11 node transmit the binary exponential each node wait for it term and get back-off until the transmission is in progress. Malicious node keep the wireless medium busy and start loaded frames transmitting which tend to capture the wireless channel by sending data nonstop as result neighbours nodes back-off endlessly. Therefore other nodes find the wireless medium busy and keep back-off so the performance of the network degrades this way.

Virtual Jamming attack 802.11 In virtual jamming attack malicious node sent RTS packets continuously continuously on the transmission with unlimited period of time. During this entire process malicious node effectively jam the trans transmi missi ssion on with with a large large segme segment nt of trans transmi missi ssion on on the wire wireles lesss channe channell with with small small expenditure of power. This attack is much effective than physical layer jamming as this attack consume less battery power compare to the other physical layer jamming attack. For example node M is a malicious node and it starting sending a false RTS packet to node R with a large frame. When nodes G and H receive packet on wireless channel they both become blocked for a certain amount of time as apply for node M as figure show in 3.1. [4]

BUSYRTS

Name: Sheraz Salim BUSY

BUSY

Page 13

Master Project CO4804 BUSY

M

CTS R

G

H

Figure 3.1 Virtual Jamming attack 802.11

2.2.3 Traffic Monitoring and Analysis Attack These Th ese attac attacks ks are to ident identify ify the charac characte teri risti stics cs of commu communic nicati ation on on radio radio wirel wireless ess transmission which could lead it for future useful information for malicious attacker to launch attack against the targeted network. This attack identify attack identify the communication nodes and their physical location, information like the amount of data transmitted, number of nodes in the network, channel frequency and spread spectrum etc. WLAN, cellular network and even satellite network suffer from these vulnerabilities. These attacks are not considerable for MANET but they are fall into other WLAN attacks

Single adversary attack (SAA) In this attack malicious node intrude into an authorized target network and starting generating IEEE 802.11 massive flow signals to authorized nodes on the target network, the goal of malicious node is to drain the battery power of authorized nodes and to reduce the available channel capacity of authorized nodes on network. This attack is known as single adversary attack. [5].

Colluding adversaries attack (CAA) In this attack two or more malicious nodes involve having understanding to contribute using 802.11 signal and start generating the signals to each other. Hence the channel get busy for

Name: Sheraz Salim

Page 14


other nodes therefore the targeted nodes in the network cannot communication communication to each other. This attack is called colluding adversaries attack [5].

2.3 Network Layer Attack Network layer is the third lowest layer of OSI reference model. The function of network layer in OSI OSI laye layerr model model is to provi provide de the the servic services es for exchan exchangin ging g the the indiv individu idual al piece piece of data/information over the network between identified end devices. To achieve the integrity four basic processes are involved in it. [6] •

Addressing

•

Encapsulation

•

Routing

•

De-capsulation

The network layer in MANET uses ad hoc routing and does packet forwarding. In MANET nodes act as host and router. Therefore router discovery and router maintains in the MANET is effectively concern. Thus attacking on MANET routing protocol not only disrupt the communication on the network even worst it paralyzed the whole communication all over the network. Therefore, a security in network layer plays a vital role to ensure the secure data communication in the network. To evaluate the security threats in network layers, threats in network layer are classified as: •

Routing protocols

•

Attacks on network layer.

2.3.1 Routing Protocol Routing is process of moving packets from source network to a destination network in internetworking devices. Routing protocols use a routing algorithm which is a mathematical formula to forward the packet to its destination. The routing algorithm varies from one rout routin ing g algo algori rith thm m to anot anothe herr etc etc OSPF, SPF, BGP. BGP. In wire wirele less ss netw networ ork k two two type typess of communication communication paradigm exists, which use different routing algorithm algorithm due to the nature of it use, infrastructure uses the traditional wired routing and in Mobile Ad hoc wireless network nodes are connected by wireless link through physical layer and data link layer, thus the network topology may keep changing randomly due to the movement of nodes. Routing protocols that are used in traditional wired network finding a path to route the data flow, for example packets from a source node to destination node, cannot be applied directly due to the natur naturee of Ad hoc wirel wireless ess netw network ork,, as frequ frequen entt path path breaks breaks durin during g ong ongoi oing ng sessi sessions ons,, Name: Sheraz Salim

Page 15


disruption commonly occur during transmission/receiving. Situation like them are not reliable links links in traditi traditional onal wired wired network network where where all node are fixed. fixed. Tradit Traditiona ionall wired wired network network protocol find path easily when path break. Convergence is slow in wired networks. Therefore, routing protocols for ad hoc network must be able to perform effective and efficient mobility as they are dynamic and no centralized infrastructure present. Many routing protocols have been proposed in MANET. The aims of these routing protocols are to provide a reliable and secure communication and remove fault in existing protocols.

Proactive (Table Driven Routing Protocols) Proactive protocols maintain up-to-date routing table information in their routing tables and regularly regularly update the changes in the topologies as appears. When a topology change occurred occurred nodes on the network broadcast the update routing table throughout the whole network to maintain up-to-date routing information. They work similar to the traditional wired network routing protocols i.e. RIP, RIP2. The protocols from this class are not feasible for large number of network as this cause more overhead in their routing tables and consume more bandwidth. The examples of different varieties of table driven protocols are given below: •

Topology Dissemination Based on Reverse Path Forwarding (TBRPF)

•

Destination Sequenced Distance vector routing (DSDV)

•

Optimised Link State Routing Protocols (OLSR)

•

Cluster Gateway switch Routing Protocols (CGSR)

•

Fish eye State Routing Protocol (FSR)

•

Wireless Routing Protocol (WRP)

Reactive (On- Demand Routing Protocols) Reacti Reactive ve protocol protocolss are also known as On-Dem On-Demand and Routing Routing Protocols, Protocols, their function function is entir entirel ely y differ different ent than than proact proactiv ivee routi routing ng protoc protocols ols.. Th They ey do not main mainta tain in the routi routing ng information of the network topology; they get necessary information of the topology when they are required. Additionally these protocols do not share routing information frequently. The examples of different varieties of proactive protocols are given below: •

Dynamic Source Routing Protocol (DSR)

Name: Sheraz Salim

Page 16


•

Ad Hoc On Demand Distance Vector Routing Protocol (AODV)

•

Temporally Ordered Routing Algorithm (TORA)

•

Associativity Based Routing (ABR)

•

Location aided Routing (LAR)

Other Routing Protocol Routing protocols are backbone of the network; a number of routing protocols are invented according according to network requirements and according according to their hierarchical routing. Proactive Proactive and reactive protocols are mostly discussed in MANET there are two other routing classes of protocol called Hybrid and Hierarchical. The Hybrid routing protocol is a combination of reactive reactive and proactive protocols. The Hierarchical routing protocols used for scalable routing sche scheme me and and set set up a hier hierar arch chy y whic which h is fall fallow ow in way way of anti anti trai trail. l. Th Thee exam exampl plee of Hierarchical routing protocol is;

Com Compari pariso son n

•

Hazy Sighted Link State HSR

•

Zone Routing Protocol ZRP

•

Distributed Dynamic Routing Algorithm DDR

of Proa Proact ctiv ive, e, React eactiv ive e

and and

Hybr Hybrid id Ro Rout utin ing g

Protocol in MANET Routing Protocol

Proactive

Reactive

Advantages

Disadvantage

Rapid establishment of

Convergence time is slow,

routes and less delay.

resources amount is used

Routing information is up-

heavily. Routing information

to-data periodically.

flooded in whole network.

Obtain the required path

Routes are not up-to-date,

when necessary, saving

large delay.

resources, do not exchange routing table periodically and Loop free. Hybrid

Up-to-data routing

Required more resources for

information, limited search

larger size zones. Arbitrary

cost and scalable

proactive scheme within

Name: Sheraz Salim

Page 17


zones. Table 4.Comparison of Routing Protocols [7]

Attack on Network Layer In network layer, the attackers aim is to disturb the network traffic by attacking on network layer, attacker attacker injects a malicious packet between the source and the destination destination nodes trying to get control of the network. Once the network is hijack attacker can easily misguide the network and absorb the traffic. To understand this situation we shall look at the example as given given below, below, Figure Figure 2.1 [8], [8], network network of three three nodes nodes S, X and and D are are forme formed d nod nodee M is malicious node and it trying to broadcast false RREQ packet to node X to get control of the network. Figure Figure 2.2 [8] shows the malicious node successfully gets into a network with false RREQ. This example shows with having no route security between nodes therefore any malicious node disturb the traffic on MANET.

S

D

X

M Figure 2.1 MANET malicious node access attack on network

S

Name: Sheraz Salim

X

M

Page 18


D

Figure 2.2 MANET malicious node attack on network

Yang proposed that two kind of attack fall in network layer, routing attacks and packet forwarding attacks. [8]

Routing attacks There are different numbers of attacks in MANET routing that can be launch against the legitimate network by employing malicious nodes. The purpose of these attacks is to spoil the current routing tables that has been built up by intermediate nodes and which are handling smoothly around the network topologies. Following are the few of them which are commonly launched to trigger the routing protocols.

Routing table overflow attack In routi routing ng tabl tablee overf overflow low attac attack k vulner vulnerabl ablee nod nodee or group group of nod nodes es adver advertis tisee route route to p part artic icula ularr non nonee exist existin ing g nod node, e, which which does does not exist exist at all, all, the aim is to empha emphasi siss the legitimate nodes on the network to join the advertise route. The goal of this attack is to originate overflow of routing table that creates a large amount of routes by vulnerable node to stop creating new routes from genuine node in the network. Proactive protocol is easy to attack as proactive protocol use discover routing information before they send route to other node on the network. [9]

Routing table poisoning attack Routing table poisoning attack is classified as internal attack, as selfish node or set of misbehaving node implement this attack for purpose to save the battery life or exploit the routing. In MANET routing protocol maintain the information regarding the routes they learn of differen differentt network network topologi topologies. es. Misbeha Misbehaving ving or selfish selfish nodes nodes create create fabrica fabricated ted routing routing updates and modifies the original routing route and re-broadcasting the modified original packet to other nodes on the network. As a result network becomes inaccessible to use. [10]

Route cache poisoning attack Mali Malici ciou ouss node nodess use use the the adva advant ntag agee of rout routee info inform rmat atio ion n usin using g tech techni niqu quee know known n as promiscuous promiscuous mode where node node can hear all the raw packet packet transmitted transmitted on the network and can easily overhearing the packet and change the routing information enclose in the packet header route cache. Route cache poisoning are mostly used in on demand routing protocols,

Name: Sheraz Salim

Page 19


as each node had a cache route that has the information regarding regarding the route it has been routed to know nodes in the network, this attack is much similar to eavesdropping attack.

Packet forwarding attacks or Advance attacks Many new attacks on network layer have been identified which are brutal against legitimate network and complicated to understand. Many new advance techniques are used to attack on networks. In these attack many routing protocol for MANET are proposed to enhance to solve the problem. Therefore, it is area of interest for researcher to find the loop hole for MANET security.

Black hole attack The black hole word is driven from the physics phenomenon, which mean swallow. In term of MANET black hole means swallow the data packets. The black hole uses the concept of generating false route and dropping data. The black hole performed two steps to accomplish this. First black hole malicious node advertises itself having a valid fresh routing paths etc. shortest and stable path to the destination node. Secondly, once black hole node attract the traff traffic ic towar toward d itsel itselff and and when when the the flow flow of traff traffic ic come come to nod nodee its its start start dropping dropping the the forwarding forwarding packet and do not forward the packet as result all packet through malicious node is lost. As shown in the figure 2.3 [11]

Wormhole Attack Wormhole attack is similar to black hole attack both attack share the similar phenomena, only the difference in wormhole work with a collision with other nodes. The goal of wormhole attack is to affect the MANET routing protocols etc AODV and DSR route discovery. First attack record the packet at one location and tunnels it to another location by employing the techniques using wired network or using an off link wireless channel at different radio frequency. The tunnel between these two colluding is called wormhole.

Classification of wormhole attack Wormhole can be classified into further five categories as proposed [12], [13] •

Wormhole using Encapsulation

•

Wormhole using out of band channel

•

Open wormhole attack

•

Closed wormhole attack

Name: Sheraz Salim

Page 20


•

Half open wormhole attack

•

Wormhole with high power transmission

Wormhole using Encapsulation When nodes S broadcast for the RREQ to its neighbours node C and node D, where as node A1 and node A2 are colluding attacker. Nodes A1 records the RREQ request as forward from node C. Node A1 tunnels the RREQ to its partner A2, and rebroadcast to its neighbours H. The request is transmitted transmitted quicker than the request from node S to node A1. As a result node D decide a route D-H-C-S and delete the route it had it before in its routing table. On the other side node S choice route S-H-D which pass through A1 and A2. As shown in the figure 2.4

B

C

M

A

E D

RREQ

A, B, C, D, E Node

Data RREQ

Malicious Node (M)

Figure 2.3 Black hole attack in MANET

Name: Sheraz Salim

Page 21


Wormhole using out of band channel This stage the attack used a band channel which is in different frequency as compare to the target network. The goal is to accomplish using a wired link or using out of band high bandwidth channel between malicious nodes wireless link using a long range directional antenna. antenna. When When node S send a RREQ RREQ to node C and D neighbou neighbours, rs, node X1 and X2 are malicious nodes and using a out of band channel between them. Malicious node X1 forward RREQ to malicious node X2, there X2 rebroadcast the packet to next neighbours so S get three route request X1-X2-G compare to other route B-C-E-F-G where as the first route is shorter than the second. [14]

X 1

A

X 2

Attacker1

I

Attacker 2

D

S

Target

B

C

E

F

RREQ RREQ through worm hole RREP Wireless link Wormhole link Figure 2.4 Wormhole attack Redraw [Rashid et al, 2008]

Name: Sheraz Salim

Page 22


G

Open wormhole attack In this attack malicious node keep examine the wireless medium to process the discovering RREQ packets, in the presence of malicious node in the network other node on the network suppose that malicious node are present on path and they are their direct neighbours.

Closed wormhole attack The attacker does not modify the capture packet nor did it modify the packet field head. The attacker attacker take the advantage when the packets are in the process to find a route know as route discovery. At route discovery process attack tunnel the packet from one side of the network to another side of the network and re-broadcast packets.

Half open wormhole attack In this attack only one side of the packet is modify from the malicious node and the other side of the malicious node do not modify the packet subsequently route discovery procedure.

Wormhole with high power transmission In this attack malicious node use maximum level of energy transmission to broadcast a packet, When malicious node received a Route Request (RREQ) by using route discovery process, it broadcast the Route Request (RREQ) at a maximum level of energy of it power so the other node on the network which are on the normal power transmission and lack of high power capability hears the maximum energy power broadcast they rebroadcast the packet towards the destination. By doing this malicious node get more chances to create a route between source and destination without using colluding node.

Jellyfish attack Jellyfish Jellyfish attack is much similar to black hole attack, Jellyfish Jellyfish attack use the same techniques techniques as black hole does. Jellyfish first invade on the forwarding packet once it get hold of the packet it start delaying the data packet for a certain amount of time before forwarding them. By doing this network performance significantly end to end delay and delay jitter occurred. In jellyfish attack malicious node send/receive route request and route reply normally. The main mechanism of jellyfish is to delay packet without any reason. [15]

Gray holes attacks Gray Gray holes holes attacks attacks have two charact characteris eristics tics,, first first a malici malicious ous node broadca broadcastin sting g falsely falsely having a valid route to destination node with goal of intercepting intercepting packets. Secondly the node drops the intercepted packets. [16] Claims malicious malicious behaviour in many means. It could drop

Name: Sheraz Salim

Page 23


coming packet from destination while forwarding forwarding packets for other nodes on network. Other different approach Gray holes malicious node can do is to drop all forwarding packet for some time may behave malicious but switch to normal behaviour later.

Byzantine attack These are vulnerable Byzantine attacks as set of few or numerous node work as cooperation to launched the attack against the target network. The compromised intermediate nodes or set of nodes worked as collusion collusion to form an attack. The compromised nodes create routing loops and may drop packets. Byzantine attacks are hard to detect.

Packet Replication In this attack malicious node generate the replicates stale packets and start sending it to a trigger network, in result the nodes on the network get overload of these packets and it will consume the bandwidth of the network and start consuming the battery of the nodes as well.

Rushing attack The authorized node in on-demand routing protocol require a RREQ packet in attempt to find a path to destination. When a malicious node receives a RREQ packet from a source node it rapidly broadcast it throughout the network topology before the other nodes on the network topology receives RREQ packets. When nodes on the network received the original packet they assume those packet are duplicate as they already received the packet form the malicious node and discard the original pack

2.4 Transport layer Attack Transport layer is the fourth layer of OSI layer model. MANET Transport layer can be categorized categorized in to connection oriented and connection less protocols which maintains maintains the endto-end flow connection, end-to-end delivery and congestion control of communication among nodes on the network. Transport layers DoS attacks are vulnerable as TCP does not have mechanism to differentiate whether it loss data was a congestion error or malicious attack. Fallowing attacks are classified on this layer. Transport layer attacks; which are as fallow [17].

Replaying acknowledgement Replaying Replaying attacks are more vulnerable than any other attack on transport layer. The MANET transport layer protocols TCP- Reno are vulnerable as the function of this protocol is to acknowledgement of the same segment multiple times signify negative acknowledgment. Name: Sheraz Salim

Page 24


Malicious nodes take the advantage and start replying acknowledgment multiple times to make the source node consider that packets are not transmitting successfully.

SYN flooding attack In this attack the malicious node sent a large amount of TCP connection is created via a victim node by only opening a half window through. Due to nature of this attack malicious node never open the full connection connection to handshake. handshake. As like traditional traditional network do, two nodes must establish TCP connection using three way handshakes to process the communication. Once the attack is launch the malicious malicious node generating vast amount of SYN data packets to faulty node.

Figure 2.5 SYN flooding attack Hijacking [18]

2.4.3 Session hijacking In session hijacking the malicious node spoofs the IP address of the victim and concludes the right sequence number and implements implements a DoS attack. Session hijacking giving a opportunity opportunity to a malicious malicious node to act as a authorized node. Once the DoS attack is performed performed the target nodes become occupied for time being. The malicious node masquerades as one of end nodes of the session and hijack the session.

Figure 2.6 Session Hijacking [18]

2.5 Application layer Attacks Application layer is upper last OSI layer the function of this layer is to provide the end to end communication between nodes; application layer represents the software application that provides provides the network services and connection connection to lower layer of OSI model. Many application

Name: Sheraz Salim

Page 25


layer attacks has identified in MANET as traditional wired network and MANET share the elements and parameters to function and to perform the application interference to share the same similarities. As MANET is self-organised nodes therefore they are mostly like to get attack from malicious nodes.

2.5.1 Application layer attacks MANET The application layer interacts with software application of the OSI layers closest to end user. The functionality of application layer is to identifying communication protocols like HTTP, FTP, SMTP and TELNET etc and maintaining maintaining by direction direction flow of communication communication between end nodes on the network. Application layer faced vulnerable activities like other OSI layers does. However the attacks launch attack on network through malicious malicious code and repudiation attack because this layer supports varieties of protocol. Once get attacked node not only suffer from degrading the network it also effect on it Operating System (OS), Trojan and Malware Malware are the common examples examples of applicat application ion layer layer attacks attacks.. An attack attack can launch launch application Denial-of-Service attack (DoS), it could be accomplished by attack start sending a large amount of request to application layer. Like HTTP attack flood hundreds of page request to web server which bring down the capability of the processing time. This process highly build computational computational load on MANET server and may degradation the network of loss of availability of the application.

Malicious attack In this attack a malicious malicious node attack on the operation system (OS) of the other host node on the network. Malicious node sent a Malware or Trojan virus in to a victim node. These attacks are severe and cause victim node to stop using the network resources. Malicious Malicious code attacks attacks are a vulnerab vulnerable le program programme me that that attack attack on host nodes. Trojan Trojan virus virus infects infects the operation system however Malware code copy itself on the network and keep spreading around the network. the aim of these attack are to constrain itself on the network and whenever a new nodes get IP access it start itself on the operation system on the node without givin giving g any notice notice to nod node. e. Th These ese atta attacks cks mostl mostly y seek seek the speci specifi ficc infor informa mati tion on on the legitimate node and sent information to malicious node which will be used to collection personal information and specifics information to attack on other nodes.

Repudiation attack Transport layer and network layer security is not enough to prevent the attacker to attack on the nodes in network. Repudiation is the attack which is by passed by a attacker from transpor transportt and network network layer. layer. Repudiat Repudiation ion attacks attacks refer refer as denial denial of partici participati pation on in the Name: Sheraz Salim

Page 26


communication. Repudiation attack can be seen as Malware where as an attacker node keep accessing the system as a selfish node and deny any conducting operation which is coming from system in order to communication on the network.

2.6 Cross layer security Attack Attack on MANET exist on number of OSI reference model, but many of few attacks can be launched are not specifically according to OSI reference model they are known as cross layer attacks. However MANET attacks are distinguish in to active and passive attacks. Active and passive give a better understanding of attack which is implemented from malicious node to legitimate networks.

Active attack In active attack the packet transmitted over the wireless radio frequency could be modified or corrupt during the data exchanged among network nodes from intercepting of malicious node to trigger network. A malicious node can modify the packet by setting up false information into packet header, as result routing events puzzled and degrades the network performance, Active attacks can be classified in two categories internal attack and external attack

Internal attack Internal attack are took place inside the network, selfish nodes misbehave against the network accordance accordance to be faithful to network; the purpose of selfish node is to save the battery life or its own purpose and not forwarding packet to other nodes. These attacks on network are hard to detect compare of external attacks as attacker is inside of the network and damage the resource of network by not cooperation.

External attack These attacks are take place from outside the network from those those node that do not are part of the legitimate network. To prevent these attacks as a caution to use an encryption technique mechanism and using firewall to block that access of unauthorised node from network resources.

Passive attack The function of the passive attack is to listening listening the radio frequency channel channel and get as much as information on the network traffic before launch any attack whist the passive attacks are laun launch ched ed to stea steall the the impo import rtan antt info inform rmat atio ion n rega regard rdin ing g to targ target et netw networ ork k such such as eavesdropping and traffic analysis. These attacks are also harder to detect.

Name: Sheraz Salim

Page 27


i)

Modification attack

In this this attack attack the malicio malicious us node gather the requirem requirement ent informa informatio tion n on which which routing routing protocol network is running this can be accomplish by using eavesdropping to hear the raw packet transmission. Once the required information is gather malicious node create changes in the routing table of the target network and start broadcasting it. Due to the nature of the MANET they are mobile nodes new path are discovered and break frequently. Malicious nodes take the advantage of this phenomenon and participate in packet forwarding process and launched the message modification attack. [25] ii)

Man-in-middle attack

This attack is also known as Tunnel attack, the attacker take control of the transmission between the source and destination by intercepting them. The intension of attack is to either to change sent messages or gain the useful information on the intermediate nodes on network. iii)

Fabrication attack

Fabrication attack is also called as tampering attack, in this attack malicious node do not interrupting or modifying any routing table thus the attacker fabricate its own packets and transmit it on the network to create a chaos to bring down the network. Fabrication attacks can also be launched from the internal misbehaving nodes like route salvaging attacks. iv)

Route salvaging attacks

This attack is launched by internal nodes such as misbehaving nodes the purpose of this attack is misbehaving node salvage their own packets duplicate them and rebroadcast them on the network with no error messages. messages. This attack could be more dangerous for the network if this attack is implementing with the collaboration of other misbehaving nodes. These attacks not only consume unnecessary bandwidth bandwidth but also drain off the resources of the other intermediate and destination nodes. v)

Sleep deprivation attacks

The aim of this attack is to drain off limited resources in MANET nodes etc, energy power, b by y keep keep sendi sending ng the route route reques requestt to the trigg trigger er nod nodes. es. Hence Hence routi routing ng protoc protocol, ol, sleep sleep deprivation attack can be launched by flooding the unnecessary routing packets in order to make the node unable to participate in the network and unreachable by other nodes on the network. vi)

Location Disclosure

Name: Sheraz Salim

Page 28


A location disclosure attack can expose information about the network topology and the location of nodes and structure which are set place. The information of location disclosure may be very helpful in regards to adjacent to the target or physical location. Routing hop limit is used to measure the distance and sending ICMP error is recorded. Using these techniques attack can find out the target nodes. vii)

Information Information Discloser attacks

In this attack a node can disclose the confidential information to malicious node, like location or structure of the network topology which is useful for the malicious malicious node to launch a attack on a specific network in result malicious node gather all the information which it need and which node is suitable to implement a attack.

Chapter NO.3 3.1 MANET Security Services Security in MANET has been an active research area from past few years. MANET posed several security security threats and challenges challenges when it comes to security. Security Security is a critical issue as MANET are dynamically move able nodes, security is a big challenge to make the commu communic nicat atio ion n relia reliable ble and and secure secure from from eaves eavesdro dropp pping ing as wirel wireless ess mediu medium m is more more susceptible susceptible to attack. In respect to MANET security the goal is to provide a ultimate security among nodes to communicate such as authentication, confidentiality, integrity, and nonrepudiation are the basic requirement of the secure communication on the radio frequency medium.

Authentication Authentication is the mechanism that node use to validate the data when transmission is taking taking place place by authoriz authorized ed nodes, nodes, this process of commun communicat ication ion is authenti authentic. c. Without Without authentication malicious nodes get access on the network and data can be modify without any Name: Sheraz Salim

Page 29


prior notice to authorized nodes. Authentication stop the access of other malicious node to function or to duplicating data to assure the recipient that message is from that source node sent to other node in network.

Confidentiality Confidentiality Confidentiality purpose is to ensure the access to legitimate nodes on data, as confidentiality could be compromised by other means through malicious nodes. MANET data that send to other nodes have to be comprehensible to receiving node only, there is a probability that message can be intercepted by intruder node. In mobile ad hoc network confidentiality is vital part for securing the transmission on wireless radio as sensitive of data. As an example the transmitted messages send on mobile ad hoc network can be caught by adversary appearing as authenticated mobile node by using a spoofing attack, as result this vulnerable activity malicious node get access to that data. To avoid this problem is used a technique called data encryption known as cryptography.

Integrity Integrity ensures the process of changing or to interception of modifying the information by authorized nodes in MANET other task of integrity is to make sure the message transmitted on the wireless medium never corrupted, a massage can be corrupted by malicious nodes. By doing this kind of attacks the adversary may change or duplicate the data packet information information message or could redirect the network traffic to other routes or to a different destination node.

Availability Availability proceeds the survivability and functioning of network to provide guaranteed services at all the times despite malicious or misbehaviour nodes on the network. If a malicious node or denial-of-service (DoS) attacks launch against the network, where DoS attack can be launched at any layer of OSI model on MANET. A malicious node use jammer to obstruct the transmission on physical layer, or data link layer to disable the back off mech mechani anism sm or on netwo network rk layer layer to halt halt routin routing g by distu disturbi rbing ng route route disco discove very ry proce process. ss. Additionally, malicious node brings down network resources. Therefore to combat these challenges key management is fundamental for implementation of any security framework.

Non-Repudiation Non Non repudiat repudiation ion ensures ensures the process process of sending sending informa information tion among among nodes nodes on wirele wireless ss medium by sending nodes and receiving nodes. The nodes does not deny of information that be transmitted or received by using a special mechanism called digital signature know as cryptography which functions unique identifiers for each other. [19]

Name: Sheraz Salim

Page 30


Characteristic

Aim

Remarks

Confi onfide dent ntia iali lity ty

Preve revent nt the the inf infor orm matio ation n to to

Data encryption

unauthorized access Integrity

avoid illegitimate deletion

Hash function

modification Availability

Network resources

Tackling MANET

accessible to nodes within the network Non Non-r -rep epud udia iati tion on

Stop Stoppi ping ng mal malic icio ious us node nodess

Using of signature on

to access the network

packets/messages

Table 5. A Review of Security Parameters in Mobile Ad-hoc networks. [26]

CHAPTER NO.4 IMPLEMENTATION 4.1 Discrete Event Simulator (DES)

Discrete event simulator (DES) is a modelling of system which defines inconsistent value changes on the condition of discrete set of point in time. Discrete event simulator is mostly use for a packet by packet model of the network topology in network communication. The simulation models are measured or analysed by numbers to a certain extent of analytical methods. Analytical simulator which is combines of mathematical formulas etc, algebraic equation, stochastic process or differential equations to solve the model requirements [20] [21]. There are different types of simulator that can distinguish etc, protocol, processing. 4.2 Continuous simulation

Continuous simulation is the system performance which can be characterized by different equations and the simulation consists in solving the equation. It is a simulation of a system of liner or non-liner differential equations. This set of equations could be represented in time domain or transform domain. [22] 4.3 Test bed Labs

The test bed are use for experimental purposes in labs where real systems are placed in to get the outcome of the selected protocol or application to measure the performance for future purpose or analyse the traffic point to point delay. The implementation of selective protocol

Name: Sheraz Salim

Page 31


and applications are harder to implement as they have their own drawback like the cost of the hardware to install is expensive in terms of use and to monitor these systems [23]. 4.5 Simulation tool

To demonstrate a research on MANET security issues there are different variety discrete event event simulat simulator or availabl available. e. Such Such as NS-2[27 NS-2[27], ], OPNET[ OPNET[28], 28], Glomosim[29], Glomosim[29], QualNet[30], QualNet[30], OMNet++[31] . I decide to work on discrete event simulator OPNET MODELER 14.5, the reason for using this software is because of it highly rich features, it has a leading atmosphere for network network design and modelli modelling ng simulat simulation ion,, it support support all most all industry industry standards standards network protocols, applications and devices. OPNET Modeler generally produces accurate result compare to other discrete simulators simulators [24]. OPNET is recognized largely from many IT organizations. In addition to that, it’s programming library help researchers and professionals to modify the network component to measure the performance in the OPNET simulation. The OPNET environment includes the study of all phases etc, Model design, data collection, simulation and analysis. 4.6 The Architecture of OPNET

Thee metho Th methodol dology ogy and the conce concept pt whic which h make make the OPNE OPNET T user user frie friendl ndly y is it simple simple architecture. Highly rich modelling environment for modelling and perform the evaluation of communication and distributed system are the most appealing features. The OPNET software suits come with a number of tools, each single suite emphasis on particular characteristic of modelling task. These tools are classified in to three parts as shown in Figure 9.2, all these three parts are inter-connected to make to flow of simulation data to generate. These three phases of modelling and simulation projects are

1). Specification 2). Data collection and simulation 3). Analysis

Re-Specification

Initial Specification

Name: Sheraz Salim

Data Collection and Simulation

Page 32

Master Project CO4804 Analysis

Resource: OPNET MODELER Documentation The phases are accepted to execute in sequence path and form and simulation cycle.

4.7 Intelligent Pulse Jammer Attack Model In this thesis we have implemented the intelligent pulse jamming attack, as jamming devices can be easily implemented to attack on the network. The function of Jammer is to deny network network transmis transmission sion services services to authoriz authorized ed users users by generat generating ing noise noise on the wireles wirelesss medium in order to block the access for authorized nodes. Jammer node keep sending the high frequency packet on the wireless medium at constant rate or at pulse rate by prohibit the traffic flow to authorized nodes on the wireless medium. Intelligent pulse jammer attack is created and implemented on “Ad Hoc On Demand Distance Vector, Geographical Routing Protocol, and Optimized Link State Routing”, all these routing protocols are implemented on each each single single netwo network rk scenar scenario io wher wheree first first norma normall traff traffic ic is genera generate ted d by defin defining ing the the application and profile parameters later by introducing the intelligent jammer attack on the network to compare the results. There are many different kind of jammer which can be use against the target network; most commonl commonly y are Constan Constantt Jammer Jammer,, Decepti Deceptive ve Jammer Jammer and Random Random Jammer Jammer and Reacti Reactive ve Jammer. Scenario 1, 2, and 3 focus on intelligent pulse jammer attack, the reason to call jammer a intelligent is because it’s pulse off time and pulse on time are the main parameter whic which h act act on jamm jammer er to beha behave ve on and and off off at cert certai ain n time time as defi define ne to gene genera rate te the the transmission. Additionally the jammer node transmit power is set lower than the normal nodes on the network, is to prove that jammer with low transmit power can create a great impact by reducing the throughput of the target network and calling our pulse jammer as intelligent pulse jammer.

4.8 Misbehavior Node Attack Model The purpose of misbehavior node is not to function properly in the network. Misbehavior nodes achieve its goal by acting maliciously. It stops forwarding packet to other nodes, by simply start dropping the packets, consume the bandwidth of the network by broadcasting route when not necessary etc. The misbehavior nodes stop performing the basic task and as result the network become congested congested and the traffic on the network lead to delay of data and degrade the performances performances of the network. In this thesis we examine the misbehavior node on

Name: Sheraz Salim

Page 33


Optimized Link State, Geographical Routing Protocol, and Ad Hoc On Demand Distance Vector routing protocol. protocol. We have make changes in misbehaviour nodes to act maliciously maliciously by dropping the packet which it received from other node on the network and make delay to forward the packet to it other node on the network. To act differently we change the packet size and packet inter-arrival time so the misbehavior nodes behave different rest of the network. For each single scenario first normal network traffic is generated and later by introducing three misbehavior nodes on each single scenario and comparing the results.

4.9 MANET Network Model A campus network scenario 800x800 is created, 29 MANETs nodes are deployed on OPNET Modeler simulator. We use IEEE 802.11 network standards for MANET nodes. The numbers of nodes in MANET MANET scenario are kept as 29, the simulation simulation run time for each scenario is set at 300 seconds with the seed value 128 and simulation kernel is set as optimized mode. The number of network scenario are six, each network scenario is implemented with different routing protocol. First three scenario are on Intelligent Intelligent pulse Jammer attack, Scenario 1 focus on intelli intelligent gent pulse pulse jammer jammer attack attack on Optimiz Optimized ed Link Link State State Routing Routing protocol protocol (OLSR), (OLSR), Scenario 2 is on intelligent pulse jammer attack on Geographical Routing Protocol (GRP), scenario 3 is on intelligent intelligent jammer attack on Ad-Hoc On Demand Distance Distance Vector(AODV) Vector(AODV).. Othe Otherr thre threee netw networ ork k scen scenar ario io are are on misb misbeh ehav avio iorr atta attack cks, s, scen scenar ario io 4 exam examin ines es the the misbeha misbehaviou viourr attack attack on Optimi Optimized zed Link Link State State Routing Routing protocol protocol (OLSR) (OLSR),, scenario scenario 5 on Geographical Routing Protocol (GRP) and last scenario 6 describe the misbehaviour attack on Ad-Hoc On Demand Distance Vector (AODV).

4.10 Application Configuration Setting Application configuration is use to run the applications on MANET nodes on the network. The function function of applicat application ion configur configuratio ation n is to specify specify the applicat applications ions that will will use to configure user profile. In order to run any network simulation in OPNET modeller there is a need to define the application which will generate traffic on the network. The applications we decided to run on our network are FTP (Medium Load), EMAIL (Medium Load), and DATABASE (Low Load). The following figures show the attributes of the application configuration setting.

Name: Sheraz Salim

Page 34


4.11 Profile Configuration Configuration Setting The function of profile configuration is to define the activity prototype of user in requisites of the application used over a period of time. Profile configuration configuration use the duration of time that each application runs on the network node. Profile configuration also specific the operation mode mode as serial serial (Orde (Ordered red), ), seria seriall (Rand (Random om)) and simu simult ltane aneous ous.. For For FTP, FTP, EM EMAI AIL L and and DATABASE applications the start time set as uniform (5,10) and duration set as end of Profile where as the number of repetitions repetitions are set as unlimited for the above applications. applications. All the application profile is set as serial (ordered) in profile configuration.

Name: Sheraz Salim

Page 35


4.12 Mobility Configuration Setting The mobility configuration setting allow the MANET nodes to move around in random direction so the link between nodes can break and new link will be establish by discovering new routing tables. Therefore, I have made some changes in Mobility random waypoint parameters. Speed is set as uniform_int (0,10), pause time is set as constant (50), start time set as constant (10) and stop time is left default end of simulation.

4.13 Node Model Traffic Packet setting In MANET nodes act as router and host to forward the packet to other nodes on network. In our simulation network scenario scenario all the MANET are mobile nodes and all nodes trajectory is selected as vector, traffic generation parameters for start time is set as 5.0, All MANET nodes Name: Sheraz Salim

Page 36


are attached with a standard wireless IEEE 802.11 network card with data rate (bps) 11Mbps, Institut Institutee of Electr Electrical ical and Electr Electronic onicss Enginee Engineers rs (IEEE) (IEEE) define define three three main main operati operational onal standards for wireless LAN i.e. IEEE 802.11 a/b/g, These all three standards belong to IEEE80 IEEE802.11 2.11 protocol protocol family family.. The most most deploye deployed d wirele wireless ss network network standard standard is 802.11b, 802.11b, Majority of the public use this standard. The data transfer rate is 11 Mbps, whereas the actual data transmission transmission rate achieved is approximately approximately 4-7 Mbps. we shall define the packet length according according to the IEEE 802.11 which is 0 to 2304 bytes plus 28 bytes header filed. To set the packet value of IEEE 802.11 in MANET node we change the packet size of MANET node to Packet size (bits) exponential 16000 bits, where 16000 bits equal to 2000 bytes. IEEE 802.11 b specification describe if network is larger than 2304 bytes the source node discard the packet. packet. The inter-a inter-arriv rrival al time time is exponent exponential ial (.03) (.03) unless unless otherwi otherwise se specifie specified. d. The RTS threshold is define 128 bytes on MANET nodes, data packet whose size is WLAN MAC header of 28 bytes exceed threshold need RTS/CTS frame exchange before transmission of the actual packet over the wireless medium. Buffer size of each single MANET node is define as 102400 rest of the value of the MANET nodes are left default etc. Channel settings, Physical Characteristics, Large packet processing.

Name: Sheraz Salim

Page 37


4.14 Creating Intelligent Pulse Jammer Node Model Creating Intelligent Pulse Jammer node model is different than create a MANET node model. The reason of that is Pulse jammer node has different structure than MANET node it has radio transmitter that continuously generate the noise on wireless medium unless specified by it source attributes to change it pulse on and pulse off setting. Pulse Jammer node can be deployed as fixed, mobile, and satellite node. To implement intelligent pulse jammer node on netwo network rk scena scenari rio o we have have to modif modify y the the Jamme Jammerr nod nodee attr attribu ibutes tes in OPNE OPNET T Mod Model eler er simulator. Jammer node band frequency is set as 2402, and the jammer bandwidth specifies the frequency in MHz as 100,000. The jammer transmitter power is set as 0.001. Once defining the attribute of the pulse Jammer node, we shall define the source attributes of pulse jammer node, pulse off time is set as 0.000399 and pulse on time is set as 1E-006.

Name: Sheraz Salim

Page 38


4.15 Creating Misbehavior Node Model Creating Misbehavior node is similar to normal MANET node. By applying the different packet setting misbehavior node act different on the network. We have change the packet size as uniform (190,220) and packet inter-arrival time is set as exponential (.005)

4.16 Scenario 1 Performance of Optimized Link State Routing (OLSR) protocol under Intelligent Pulse Jammer attack

First, we implemented Optimized link State protocol on 29 MANET nodes all node are configure default OLSR setting setting of Ad-Hoc routing parameters in OPNET Modeler. The area Name: Sheraz Salim

Page 39


is defined defined as of 800x800. 800x800. Applica Application tion configur configurati ation, on, profile profile configur configurati ation on and mobili mobility ty configuration are defined and implemented in order to generate the traffic under OLSR routing protocol. The traffic result is recorded. Secondly, we placed two intelligent pulse jammer nodes in the OLSR network within 29 MANET nodes and named the scenario as OLSR_JAMMER, the jammer nodes will transmit packets on pulse time setting on the wireless medium until simulation time end. We have recorded the results against the normal network with OLSR and with intelligent pulse jammer nodes under OLSR and compared with in term of Throughput, Throughput, Delay, Traffic Traffic sent, Traffic Traffic Received, packet packet dropped and error messages.

4.17 Scenario 2 Performance of Geographical Routing Protocol (GRP) under Intelligent Pulse Jammer attack

In this scenario, first we implemented the Geographical Routing Protocol on MANET nodes with the default GRP setting in OPNET Modeler. The area of GRP is set as 800x800 so nodes can freely move in random directions. Application Application configuration, Profile configuration configuration and mobility configuration are defined. The results are noted under the GRP routing protocol with normal traffic. Later we implemented intelligent pulse jammer attack under GRP routing protocol and named the scenario as GRP_JAMMER, the results are compared in term of throughput, delay, traffic sent, traffic received, packet dropped and error messages.

4.18 Scenario 3 Performance of Ad-Hoc On-Demand Routing (AODV) under Intelligent Pulse Jammer attack

In third scenario we use the Ad- hoc (AODV) routing protocol on 29 MANET nodes with the area 800x800. Application Application configuration, configuration, profile configuration configuration and mobility mobility are defined as in scenario 1 and 2. Once the normal traffic result are collected we duplicated the AODV scenario and name is AODV_JAMMER by introducing two pulse jammer nodes under AODV nodes. And compare the capture result against the throughput, delay, packet dropped and error messages.

Name: Sheraz Salim

Page 40


4.19 Scenario 4 Performance of Optimized Link State Routing (OLSR) protocol under Misbehavior attack

In this scenario we have implemented misbehavior node attack on Optimized Link State Routing protocol. protocol. Before implementati implementation on attack we define a network with random mobility mobility so MANET node can move around in the area of 800x800. Misbehavior node attack is different different compare to intelligent jammer attack. Misbehavior Misbehavior node act maliciously act to drop packets and stop forwarding packet to other nodes and consume bandwidth. To generate the traff traffic ic on the the netwo network rk we defin definee the appli applica catio tion n confi configur gurat ation ion whic which h conta contain in data data the application used in the network, in profile configuration different application are associated to generate the traffic. The application used in Misbehavior node are FTP (Medium Load), EMAIL (Medium Load) and Database (Low Load). In our misbehaviour node we only change the packet size of the node. Our model act genuinely because we haven’t change any power transmitter value, same RTS value be use as for normal MANET node used. Apart from packet size nothing else is modified in misbehavior node. We run a normal network with 29 MANET nodes using default setting of OLSR routing protocol within area of 800x800. Same application configuration, profile and mobility are used as mention above. All the result of normal network has noted. A new scenario is created with with the name of OLSR_M OLSR_MISB ISBEHA EHAVIO VIOU, U, here we placed placed three three misbehav misbehavior ior nodes nodes at different different position so when the traffic is generated generated among other nodes misbehavior node start dropping and stop forwarding packet to other nodes. All results are capture and compare against the normal network in terms of throughput, delay, packet delivery ratio and network load.

4.20 Scenario 5 Performance Performance of Geographical Geographical Routing Protocol (GRP) under Misbehaviour attack

Geographical Routing Protocol is implemented as default setting of GRP in OPNET Modeler under 29 nodes the area size, application application configuration, profile configuration configuration and mobility mobility is kept as mention in earlier scenarios. First normal traffic is generated under GRP routing protocol

later

duplicate

t he

scenario

with

same

parameter

with

name

of

GRP_MISBEHAVIOUR, three misbehaviour nodes are placed at different position in the area of 800x800. After simulated both normal scenario network and GRP_MISBEHAVIOUR network we compare the result of each scenario in terms of throughput, throughput, network load, packet delivery ratio and delay.

Name: Sheraz Salim

Page 41


4.21 Scenario 6 Performance Performance of Ad-Hoc On-Demand Routing (AODV) under Misbehaviour attack

The last scenario focuses on the performance of Ad Hoc On Demand Routing (AODV) under three misbehaviour nodes. The MANET nodes are configured to use default AODV routing protocol in OPNET. Application, profile and mobility defined as mention earlier scenarios. Normal traffic result capture once simulated new duplicate scenario is created with a name AODV_MISBE AODV_MISBEHAVIO HAVIOUR UR where we introduce three misbehavior nodes and once collected collected their result we compare result against the normal network.

4.22 Performance Metric and Analysis Performance metric are important aspect to perform analysis on network and to explain the implementation implementation of the simulation simulation setup, its design are explained, and a key factor to evaluate the real performance of the network. In our network simulation we decided to chose, Packet end to end delay, throughput, delay and network load. 

Packet Delivery ratio

Packe Packett deliv deliver ery y rati ratio o is number number of data data packe packets ts recei receive ved d by desti destina nati tion on nod nodes es divided by number of data packets transmitted by source nodes. 

Delay

The packet end to end delay is the average time of the packet passing through inside the network. It includes all over the delay of network like transmission time delay which occurs due to routing broadcasting, buffer queues. It also includes the time from generating packet from sender to destination and express in seconds. 

Throughput

The ratio of the total data reaches the receiver from the sender. The time it consume by the receiver to receive the last packet is called throughput [33]. Mathematically throughput can be characterized as in equation [34]. Throughput=Number of delivered packet*Packet size*8/Total duration of simulation 

Network load

Network load is the total packet sent and received across the whole network at a particular time.

Name: Sheraz Salim

Page 42




Packet Dropped

Packet dropped shows how many packets successfully sent and received across the whol wholee netw networ ork. k. It also also expl explai ains ns the the numb number er of pack packet et drop droppe ped d duri during ng the the transmission due to interference from other devices. Conclusion

This chapter describes the model of the network we implemented to focus for our results, we explain in well detail the parameters parameters chosen for application configuration, configuration, profile profile configurat configuration, ion, mobility, mobility, and we also explain explain the model model value value settings settings of MANET nodes, intelligent pulse jammer and the modification of misbehaving nodes

Name: Sheraz Salim

Page 43


CHAPTER NO.5 5.1 RESULT AND ANALYSIS In this chapter we briefly explain the several metric parameters which are packet end to end delay, throughput, network load, packet dropped. All these metric are taken from the earlier chapter scenario’s scenario’s result of each MANET simulation. simulation. All these result are first compared with their normal scenario where no malicious activities take place. We have analysis the results in term of MANET security likewise, availability, confidentiality and authentication.

5.2 SCENARIO 1 Comparison of Optimized Link State Routing (OLSR) protocol under Intelligent Pulse Jammer attack

To view view the the simu simulat latio ion n resul results ts and and to comp compare are them them we have have run run the the mode modell netw network ork simulation, a model network simulation is explained in chapter no 4. The scenario is run three times to make sure the result are accurate accurate and compared with several metric parameters. The results of the simulation are noted and are as follows:

Fig 5.1 Normal traffic without any attack and Fig 5.2 traffic under jammer attack

Name: Sheraz Salim

Page 44


The fallowing graphs shows the OLSR protocol behaviour with and without under intelligent pulse jammer attack.

Fig 5.3 5.3 Normal Normal OLSR OLSR routing routing protocol protocol

Fig 5.4 OLSR OLSR routi routing ng under intel intellig ligent ent pulse pulse

Jammer Evaluation

5.2.1 Throughput Figure 5.4 Intelligent pulse jammer attack on OLSR shows a significant result. The pulse jammers jammers reduce the throughput of the entire network by generation raw packet (noise) in the wireless wireless medium. The simulation simulation started time on both scenarios normal traffic of OLSR and OLSR with jammer attack is 5.0 second and up till 300 seconds. The overall throughput of normal network is 1,400,000 bit/sec (1.4 Megabits) however compare with intelligent pulse jammer attack the overall throughput on entire network is reduce to 1,200,000 bit/sec (1.2 Megabits). Therefore we can conclude the jammer attack use the wireless medium and decrease the network traffic throughput.

Name: Sheraz Salim

Page 45


Figure 5.4

5.2.2 Network Load The network load shows the occurred event on the network traffic by introducing the pulse jammer attack. It can be seen clearly the network delay slightly reduce when the jammer start generating raw packet on the network. Figure 5.5 shows the normal traffic on network load with average of 1,313727 bits/sec (1.3 Megabits) and later with jammer node on network showed the network load with average 1,086427 bits/sec (1.0 Megabits). The network load made a huge impact.

Figure 5.6

Name: Sheraz Salim

Page 46


5.2.3 Delay The Figure 5.7 shows the delay increased in presence of pulse jammer nodes with overall delay is start increased at the start of the simulation in pulse jammer scenario compare with the normal network. we can say that due to abnormal activities due to pulse jammer nodes on netwo network rk had make make netwo network rk more more vulner vulnerabl ablee which which refl reflect ect the the need need of confi confiden dentl tly, y, availability availability and authentication on network. The delay is increased up to 26 seconds compare to normal network at 22 seconds.

Figure 5.7

5.2.4 Packet Dropped The normal OSLR network packet dropped is lower than the OLSR under pulse jammer attack shown in Figure 5.8 the reason of that packet dropped showed that at the power transmit 0.001 still can be effective compare to normal node where power transmit is set at 0.005. 0.005. As pulse pulse jammer jammer generated generated raw packet than than normal normal OLSR OLSR network network.. The packet dropped average of pulse jammer is 1,400,000 bits/sec (1.4 Megabits) where as the normal OLSR network packet dropped average is 1,200,000 bits/sec (1.2 Megabits). If we compare Packet dropped graph of OSLR with throughput graph of OLSR we observe the throughput of OLSR normal network was 1.4 megabits but in case of packet dropped the OLSR normal network packet dropped average is 1.2 megabits. This shows the packet sent to its destination or to forwarding other nodes were successful as compare to introducing the pulse jammer reduce the performance of network in all means. Name: Sheraz Salim

Page 47


Figure 5.8

5.2.5 OLSR routing Hello packet The OLSR routing protocol is observed by implementing pulse jammer node attack on the network in Figure 5.9. The jammer attack clearly showed that it not only disrupted the throughput and network load but it also decrease the routing table of OLSR routing protocol. During the jammer attack the Hello packet decrease compare to normal network. Hello packet traffic sent rate as normal network 16,000 bit/sec (14.5 Kilobytes) and with pulse jammer is 14759 bits/sec (1.84 kilobytes) it is clearly observed during the abnormal activity cause of jammer node disturbed the routing forwarding table to other nodes due to the noise in the wireless medium.

Name: Sheraz Salim

Page 48


Figure 5.9

5.3 Scenario 2 Comparison of Geographical Routing Protocol (GRP) under Intelligent Pulse Jammer attack

Scena Scenari rio o 2 is exam examin ined ed the Intel Intelli lige gent nt pulse pulse jamm jammer er atta attack ck on Geogr Geograph aphica icall Routi Routing ng Protocol. We run the simulation and capture the results shown in Figure 5.10. The simulation run time is set as 300 seconds.

Figure 5.10 Network with GRP routing protocol the next Figure 5.11 and 5.12 show the performance GRP routing protocol without and with intelligent pulse jammer attack.

Name: Sheraz Salim

Page 49


Figure 5.11 GRP wihout Pulse Jammer attack Figure 5.12 GRP with Pulse Jammer attack

Evaluation

5.3.1 Throughput To evaluate the result of GRP routing protocol under intellignet pulse jammer attack are compare to the GRP normal network and GRP Jammer network in Figure 5.13. we analyse the throughput of the entire network with and without intellignet pulse jammer. The graph show show that that with with jamm jammer er attac attack k GRP GRP routi routing ng protoc protocol ol had a severe severe effec effectt on netw network ork throughput. throughput. The normal network throughput recorded recorded as 35,500,000 bits/sec duration of 300 seconds, where the start time to generate traffic set as 5.0 seconds. On the other hand the GRP with jammer attack showed a throughput of 1,000,000 bits/sec which is two times less than the original recorded on GRP with normal network scenario. If we convert the bits/sec rate in to megabits we get better view to understand the network performance with and without jammer jammer nodes on the network. without without jammer the total throuput throuput is 3.5 megabit per second and with two pulse jammer nodes the total throuput is 1 megabit per second.

Name: Sheraz Salim

Page 50


Figure 5.13

5.3.2 Network Load The network loads show the Figure 5.14 of GRP normal network and GRP with pulse jammer. It shows the network load increased by introducing of pulse jammer. The reason of that is pulse jammer generating a noise on radio frequency in pulse time which increases the network load on the network.

Figure 5.14

Name: Sheraz Salim

Page 51


5.3.3 Delay The delay on GRP routing protocol shows a great impact on the throughput on the network. The normal network delay is constant on x-axis however with the intelligent pulse jammer nodes the delay is increased exponentially. It has clearly showed the increase in delay will affect the reliability of the network and takes the network in to the congestion dead lock as shown in Figure 5.15

Figure 5.15

5.3.4 Packet Delivery Ratio The Packet delivery ratio conclude the traffic send and traffic received on MANETs by the destination destination nodes or forwarding forwarding data for other nodes from the source node, clearly showed showed in Figure 5.16 that pulse jammer destroyed the transmission on the radio frequency due to its malicious activity of generating noise. Analyses on graph shows the Normal network average MANET traffic sent rate is at 16,000,000 bits/sec on the other hand the normal MANET traffic traffic receive received d average average rate is 2,000,00 2,000,000 0 bits/se bits/secc with with duration duration of 300 seconds. seconds. When When implementing intelligent pulse jammer attack the MANET traffic sent and MANET traffic received rate is decline rapidly. Packet delivery ratio has highlighted the affect of pulse jammer traffic sent/received on MANET nodes. Pulse jammer node reduces the traffic sent average rate to 2,000,000 bit/sec where as the normal MANET node traffic traffic received rate was 2,000,000 2,000,000 bits/sec. On the other hand the pulse jammer jammer node severely reduce the MANET MANET traffic received rate on the network with average rate of 366,810 bits/sec (45.85 kilobytes) compare to 2,000,000 bits/sec (2 Megabits) of normal MANET traffic received average rate.

Name: Sheraz Salim

Page 52


Figure 5.16

5.4 Scenario 3 Comparison of AODV under Intelligent Pulse Jammer Attack

The result of AODV under Pulse jammer attack had been taken, implementation implementation had defined in chapter no 4. Simulations are run at number of time to make sure the results are accurate.

Name: Sheraz Salim

Page 53


We compare the results under a number of parameters. Captured Captured results of simulation simulation are as:

Figure 5.17 AODV normal traffic responses Figure 5.18 AODV under Pulse Jammer traffic

Name: Sheraz Salim

Page 54


Figure 5.18 Normal traffic response

Figure 5.19 Pulse Jammer traffic response

Evaluation

5.4.1 Throughput The normal network throughput is compared with intelligent pulse jammer attack in Figure 5.20. As we notice the difference of attack with pulse jammer lead network to congestion and decre decrease ase the the netwo network rk perfor performa mance nce.. Netwo Network rk reache reachess at lowes lowestt throu throughp ghput ut decre decrease ase at b begi eginni nning ng of the the simula simulati tion on when when jamm jammer er start start genera generate ted d the noise; noise; the the total total netw network ork throughput is 5000,000 bit/sec (5.5 megabits) whereas by implementing jammer nodes the throughput on the network is decrease. We can say the usage of channel lead to congestion and network performance is decrease

Figure 5.20

5.4.2 Network Load Network load of the whole network is shown in the Figure 5.21. Comparing the two networks, normal normal network and pulse jammer attack network shows a load with pulse jammer jammer attack decrease the network. The bandwidth available to the network is highly reduced in presence of jammer nodes. As packet sent from the MANET node to other nodes on the network are lost due to jammer nodes. This clearly reflects the availability and reliability of MANET nodes in terms of security.

Name: Sheraz Salim

Page 55


Figure 5.21

5.4.3 Delay The Figure 5.22 shows the delay between normal network and with jammer nodes. A gradual decrease on y-axis is clearly seen. By observation the pulse jammer attack the delay starting at co-ordinates of the beginning of the simulation and last it till at end. Most probably the reason of that is jammer start sending packets on the wireless frequency on it start time and it get increased due to noise in the network which prevent the MANET node to continue the transmission on the network.

Name: Sheraz Salim

Page 56


Figure 5.22

5.4.4 Retransmission Attempts Retransmission attempts occurred in network when delivery of packet is dropped or lost without reaching the destination nodes. Retransmission attempts have increased in presence of pulse jammer nodes shown in Figure 5.23. Pulse jammer network and normal network start b broa roadca dcasti sting ng reque request st for creat creatin ing g route route among among the the nod nodes es.. Th Thee norma normall netw network ork start startss bro broad adca cast stin ing g the the requ reques estt rout routee to buil build d rout routin ing g tabl tables es.. Th Ther eref efor oree the the incr increa ease se of retra retransm nsmis issio sion n attem attempts pts affec affectt direc directl tly y propor proporti tiona onall of the netwo network rk due to load load that that is increased on entire network.

Figure 5.23

5.4.5 Route Discovery Time To measure the route discovery time the Figure 5.24 showing the route discovery time of AODV under pulse jammer attack. The normal network route discovery time is constant from 7 minutes to until at the end of the simulation. The up and down route discovery time of AODV under normal network is unbalanced until the time of 7 minutes. This is because of the reason when MANET routing start initializing at the beginning of simulation it start broadcasting the routing table. Compare to under pulse jammer attack the vulnerable nodes starting the raw packet transmission in order to make the network vulnerable in terms of complex transmission or to degrade the MANET performance by making noise on radio frequency. First the jammer nodes make it complicated for MANET to performance route

Name: Sheraz Salim

Page 57


discovery time by time go it make it more worse and gradually decrease the time of route discovery.

Figure 5.24

5.5 Scenario 4 Comparison Comparison of Optimized Optimized Link State State Routing Routing (OLS (OLSR) R) protocol protocol under Misbehavio Misbehaviorr nodes

After implementing intelligent pulse jammer attack, scenario 4 focus on the result of the misbehavior misbehavior nodes, the implementing implementing and the parameter setting has already defined in earlier chapter 4. We have run the misbehavior nodes simulation network with number of time to make it show the results are accurate meanwhile compare those results with the normal network to analyse and understand the impact misbehavior nodes made on network. In this scenario, we compare the Optimized link state routing protocol under misbehavior nodes, the simulation run time is set as 300 seconds, and the fallowing graphs will give detai details ls of the the resul results ts and evalua evaluati tion on betwe between en the norma normall netw network ork traff traffic ic respon response se and misbehavior traffic response.

Name: Sheraz Salim

Page 58


Figure 5.25 & 5.26 With normal traffic response with Misbehavior nodes

Figure 5.27 & 5.28 With Misbehavior nodes traffic response in network. Evaluation

5.5.1 Throughput The Figure 5.29 shows the entire throughput of the network with misbehavior nodes and without misbehavior misbehavior nodes in the network. Throughput Throughput progressively increased to 1,5,10,313 1,5,10,313 bit/sec with in a set time of 300 seconds. On the other hand nodes on the misbehavior p perf erform orman ance ce reduc reducee the the throu throughp ghput ut of the entir entiree netwo network rk and keep keep at consta constant nt rate rate of Name: Sheraz Salim

Page 59


1,443,79 1,443,790 0 bits/se bits/sec. c. The vulnera vulnerable ble activit activities ies of misbeha misbehavior vior nodes nodes didn’t didn’t decrease decrease the throu throughp ghput ut gradu gradual ally ly but keep keep it at a consta constant nt rate. rate. Howe However ver the misbe misbehav havior ior nod nodes es throughput on the whole network is recorded at 1,390,698 bit/sec, we also observe the throughput during the end of simulation with misbehavior node is reducing to 1,390,239 bits/sec. We can conclude the throughput of the whole network will reduce more if the during of simulation time extend to more than 300 seconds. Additionally misbehavior nodes are compared with normal network traffic where the usage of wireless medium is increased at substantial level.

Figure 5.29

5.5.2 Delay The delay in Figure 5.30 show the normal traffic delay and misbehavior traffic delay. By observing the graph the network misbehavior delay started at almost the same time as with normal traffic however however the delay occurred at co-ordinates co-ordinates 12 x-axis where it can be examine examine clearly. Whereas the normal network delay at 12 x-axis and further got to at y-axis 22. On the other hand the misbehavior nodes y-axis ends at 25 y-axis. From here to till end of the simulation run time the delay is getting increased. The delay has increased systematically to higher level by introduction of misbehavior nodes on the network. The reason for this situation situation is as all nodes using the same routing protocol but due to different characteristics characteristics of misbehavior nodes act as maliciously therefore some intermediates nodes in the network follows the selected nodes to forwarding of packets.

Name: Sheraz Salim

Page 60


Figure 5.30

5.5.3 Packet delivery ratio (PDR) The Figure 5.31 shows the total traffic received and total traffic sent with and without misbehavior misbehavior nodes on the network. This process is also know is packet delivery ratio, where the total traffic sent and total traffic received packet are measure in terms of performance because it has a relation relation to real delivered delivered packet at source node and received node packets at the destination. By examine the graph we observed the rate of traffic received with having misbehavior misbehavior nodes on the network is steadily decreased. Whereas the normal traffic received co-ordinates showed the received rate at 900,000 bits/sec as passage of time of simulation process the traffic received slightly decrease and end at 600,000 bits/sec. The reason behind this is because of MANETs MANETs nodes exchange the routing table to other nodes and few packets are dropped or discarded. Compare to misbehavior nodes the traffic received gap is 200,000 bits bits/se /sec. c. As misbe misbehav havio iorr nod nodes es do not forwa forward rd packet packet to other other nod nodes es and and drop drop packet packet increased and whole network lead to dead lock in term of performance.

Name: Sheraz Salim

Page 61


Figure 5.31 The traffic sent on normal network is noted as 2,000,000 bits/sec where as the network with misbe misbehav havior ior nod nodes es are are 1,827 1,827,6 ,631. 31. We can concl conclude ude that that the traf traffic fic sent sent to netwo network rk is decreased by introducing the misbehavior nodes. As misbehavior nodes does not participate to perform it basic task for the fulfilment of the requirement of network in good means. These activities lead network to congestion and decrease it performance.

5.5.4 Network Load The Figure 5.32 represents the network load on the network with normal traffic and with misbehavior misbehavior traffic. There is a difference between normal network and misbehavior misbehavior nodes on network.

Name: Sheraz Salim

Page 62


Figure 5.32

5.6 Scenario 5 Comparison of Geographical Routing Protocol (GRP) under Misbehavior nodes

In scenario 5, network traffic is generated on Geographical Routing Protocol with normal nodes and later introducing with misbehavior nodes. All the simulation result are capture and are compared, following are the result capture with normal and with misbehavior nodes.

Figure 5.33 Normal traffic response

Name: Sheraz Salim

Figure 5.34 Misbehavior Nodes traffic response

Page 63


Figure 5.35 GRP traffic response

Figure 5.36 GRP traffic under Misbehavior nodes

Evaluation

5.6.1 Throughput To determine the network performance we first look at the throughput of the whole network with with comp compari arison son of misbe misbehav havior ior nod nodes es as shown shown in Figur Figuree 5.37. 5.37. Th Thee run time time of the the simulation is set as 300 seconds. The normal network throughput is noted as 3,500,000 bits/sec at constant rate on x-axis. Introducing the misbehavior nodes affectively reduce the throughput of the network performance at constant rate of 1,500,000 bit/sec. This showed a poor poor perfo perform rmanc ancee of netw network ork traff traffic. ic. Th Thee three three misbe misbeha havio viorr nod nodes es had decrea decreased sed the throughput of the whole network by not forwarding packets to other nodes and keep dropping the packet randomly. This concludes the performance of misbehavior nodes have poor affect on the network traffic. It also concludes the reliability of the node is essential in terms of security.

Name: Sheraz Salim

Page 64


Figure 5.37

5.6.2 Delay The delay parameters show the misbehavior nodes activities on the whole network in Figure 5.38. The delay has increased increased compared with normal network traffic. The reason of increase delay is misbehavior nodes does not cooperate with other nodes on the network. Misbehavior nodes only forward packet when they want too. Compare to normal network traffic a constant rate is noted on x-axis throughout the simulation time. However However the misbehavior misbehavior nodes delay is notice up and down due to the malicious activity on the network. It also show the packet delay time is increased to reach the packet at destination node.

Figure 5.38

Name: Sheraz Salim

Page 65


5.6.3 Network Load The network load of the normal network noted as 3,500,000 bits/sec at duration time of simulation 300 seconds in Figure 5.39. Network load shows the total load on the MANET’s node on the whole network in terms of bandwidth usage. Later implementing the three misbehavior nodes simulation result showed the whole network load is increased due to vulnerable activities of misbehavior nodes on the network. The network load reduces to 1,500 1,500,0 ,000 00 bits/ bits/sec sec.. It has concl conclude uded d that that the the misbe misbehav havior ior nod nodes es faile failed d the netwo network rk performance in every aspect.

Figure 5.39

5.6.4 Packet Delivery Ratio The Packet Delivery Ratio is compared to examine the network performance with vulnerable activities in contrast with normal network performance in Figure 5.40 & Figure 5.41. The average MANET traffic sent on the wireless medium is 1,800,000 bits/sec (1.8 Megabits) with no vulnerable activities on the network and the received traffic on normal MANET’s node is 400,000 bit/sec 0.4 (Megabits) at the start of the simulation but it has slightly increased and reached up to 600,000 bits/sec (0.6 Megabits) at the end of the simulation result. On the other hand nodes with vulnerable activities are implemented the total average of traffic received is noted 1,522,600 bit/sec(1.5 Megabits), compare to normal network

Name: Sheraz Salim

Page 66


traffic receiving packet it has decrease due to malicious nodes does not work together with normal network nodes. The average sent traffic is noted with malicious malicious node on the network at below 2,000,000 bits/sec (2 Megabits). Megabits). This clearly shows malicious malicious node are misbehavior on the network. They are not performing their duties, they are missing the packet and not forwarding the require packet to other nodes on the network.

Figure 5.40 Normal MANET Traffic sent/received Figure 5.41 MANET Traffic sent/received under Misbehavior nodes

5.7 Scenario 6 Comparison of Ad-Hoc On-Demand Routing (AODV) under Misbehavior nodes

Scenario 6 is design to examine the on AODV routing protocol under misbehavior nodes. The simulation run time is set as 300 seconds all results of the simulation are captured and are as fallow.

Name: Sheraz Salim

Page 67


Figure 5.42 Normal network traffic

Figure 5.44 AODV normal traffic

Figure 5.43 Under Misbehavior Nodes traffic

Figure 5.45 AODV Misbehavior Nodes traffic

Evaluation

5.7.1 Throughput The throughput of the whole network with normal nodes and with misbehavior nodes is simulated and results are captured in Figure 5.46. The result shows significant changes on throughput of the network with misbehavior nodes. By examine the result we observe the misbehavior nodes throughput started at the same time, when normal network nodes start.

Name: Sheraz Salim

Page 68


However later at 1 minute and 20 seconds time the misbehavior node throughput decline up until at 3 minutes the throughput of normal network nodes and with misbehavior nodes on network get equals. Form ahead of 3 minutes the misbehavior misbehavior nodes throughput throughput is increased increased compare to normal network, the throughput for normal network end of the simulation is recorded as 5,000,000 bits/sec (5 Megabits) whereas misbehavior nodes throughput is higher 5,400,000 bits/sec (5.4 Megabits). The reason of this increase throughput with misbehavior nodes on the network could be, first misbehavior nodes start dropping the packet and not forwarding forwarding the packets for other nodes on network, later later misbehavior nodes start sending the packets and forwarding packets faster than normal node in result normal nodes are not able to process the packet and they overlooked the packets.

Figure 5.46

5.7.2 Network Load The network load with misbehavior node is increased compared to normal network. The reason as explain earlier nodes doing vulnerable activities forwarding packet and sending packet more faster than normal node lead network to congestion and the performance to network is affected as shown in Figure 5.47

Name: Sheraz Salim

Page 69


Figure 5.47

5.7.3 Packet Delivery Ratio The packet delivery ratio results are analysing in terms of MANET traffic received bits/sec and MANET traffic sent bits/sec in Figure 5.48 and Figure 5.49. First we examine the MANE MANET T traff traffic ic sent sent bits/ bits/sec sec.. Th Thee averag averagee traff traffic ic sent sent with with norma normall netw network ork noted noted as 15,000,000 bit/sec co-ordination on x-axis. The starting time of the simulation was set at 5 sec, on initial start the normal traffic reached rapidly on y-axis at constant rate. When normal traffic is measured in response of malicious traffic with introducing misbehavior nodes the traffic sent is decrease heavily. The traffic with misbehavior nodes noted as 13,000,000 bits bits/se /secc during during the the simula simulati tion on time time was was set as 300 second seconds. s. Th Thee misbe misbehav havio iorr nod nodes es highlighted the nodes are acting maliciously and are not doing well in terms of cooperation with other nodes. Secondly, the MANET traffic received is noted, the average result of traffic received is recorded as 1,400,000 bits/sec, however MANET traffic received go up and down during the simulation the maximum average time of received packet is noted 2,800,000 bits/sec, later slightly decrease occurred and it start going down. The possible reason of this up and down rate of MANET traffic received is due to starting of the network nodes exchange the routing discovery, route request, and routing table among each other. When introducing the misbehavior nodes on the network the MANET traffic received is recorded lower than the normal network traffic. The average time of traffic received is 1,400,000. Thus the normal traffic received and the malicious traffic received is totally equal. However the start time with malicious nodes compare with normal nodes showed a difference of packet

Name: Sheraz Salim

Page 70


receiving. The maximum traffic received with misbehavior nodes is 2,400,000 bits/sec it’s declini declining ng as simulat simulation ion run. This conclude conclude,, the traffic traffic with malici malicious ous nodes nodes on normal normal network degrade the performance of network.

Figure 5.48 AODV traffic sent with and without Misbehavior nodes Figure 5.49 AODV traffic received with and without Misbehavior nodes.

5.8 Analysis The simulation results present the performance of OLSR routing protocol under intelligent pulse jammer attack. The comparison results taken in terms of throughput, packet delivery ratio, network load, and delay are measured under intelligent pulse jammer attack, As the pulse jammer attack model is already explained in chapter No 4. The capture results are compared with OLSR routing protocol under jammer attack. The throughput of the whole normal network with OLSR routing protocol is noted as 1,400,000 bits/sec (1.4 Megabits), the reason of that low throughput is because of the heavy applications are used on the each single MANET nodes i.e., FTP use TCP protocol, EMAIL use TCP protocol where as TCP uses three way handshake which consume a lot of network bandwidth. The comparison against the intelligent pulse jammer the throughput of the entire network with under pulse jammer attack is recorded as 1,220,000 bits/sec (1.22 Megabits). The packet delivery ratio of MANET traffic sent with normal network and with pulse jammer network is recorded as 1,891,000 bits/sec (1.8 Megabits) duration of 300 seconds. Meanwhile the normal MANET traffic receiving rate is 600,000 bits/sec (600 kilobytes) where as with intelligent pulse jammer introduction the MANET traffic received rate is 450,000 bits/sec (56.25 kilobytes). The network load for OLSR network is 1,400,000 (1.4 Megabits) and with pulse jammer

Name: Sheraz Salim

Page 71


attack the network load is decreased decreased is 1,200,000 1,200,000 (1.2 Megabits). Megabits). Here the whole throughput of normal network and the whole throughput of network with pulse jammer attack get equal to the network load. On the other side there is no considerable increased or decreased in OLSR routing routing total hello sent packets because the OLSR routing broadcast for the request of route to every node and will start the packet transmission once finding the route. It has also observed TC hello messages increased due to jammer attack. Simulation Simulation results for GRP protocol was examine under intelligent pulse jammer attack. The results were compared in terms of security considerations as explained earlier. The whole throughput of GRP nodes on normal network is capture as 3,400,000 bits/sec (3.4 Megabits), network load of the GRP nodes is 3,500,000 bits/sec (3.5 Megabits), with the implementation of intelligent pulse jammer attack the throughput and the network load with attack reduce to 1,000 1,000,0 ,000 00 bits/ bits/sec sec (1.0 (1.0 Megabi Megabits) ts),, 1,000 1,000,00 ,000 0 (1.0 (1.0 Megab Megabit its). s). We found found the the delay delay has has increased when jammer generate raw packet on the wireless radio frequency, the GRP routing traffic sent is reduce due to jammer attack as every node in network maintain the routing table at certain interval of time. GRP packet delivery ratio ratio decreased of packet when jammer nodes vulnerable the network by its malicious activities. Simulation results of AODV examine under intelligent pulse jammer attack. The results are compared in term of normal traffic parameters and with pulse jammer traffic parameters, throughput, network load, delay and packet delivery ratio is observed to analyse the results. Thee throug Th throughpu hputt with with norma normall nod nodes es is recor recorded ded as 5,000 5,000,00 ,000 0 (5 Megab Megabit its), s), where whereas as by introducing the intelligent pulse jammer nodes the throughput of entire network is recorded as 4,500,000 (4.5 Megabits), The AODV routing protocol on MANET’s nodes showed the decreased in delay, packet delivery ratio of received traffic under pulse jammer attack is decreased and for the sent traffic is almost equal to normal traffic of nodes when pulse jammer is implemented, the reason behind this could be the jammer is sending/generating packet of the wireless medium at same time compare with the normal network with AODV routing table etc, route discovery, route request etc. Hence jammer and routing table working at same time of interval. However the packet lost is compared lower in AODV under pulse jammer attack compare to other routing protocol. Simu Simula lati tion on resu result ltss exam examin ined ed for for the the perf perfor orma manc ncee of OLSR OLSR rout routin ing g prot protoc ocol ol unde under r misbe misbehav havior ior nod nodes es.. Th Thee result result showe showed d large large number number of packet packetss droppi dropping ng plac placed ed from from misbe misbehav havior ior nod nodes. es. Th Thee proce process ss of OLSR OLSR forwa forwardi rding ng packet packet MPR MPR lowe lowerr than than norma normall network as nodes with misbehavior capabilities does not cooperate on the network nodes. The throughput of OLSR under misbehavior node is 1,400,000 bits/sec (1.4 Megabits) where as Name: Sheraz Salim

Page 72


the network load is noted as same as throughput under misbehavior nodes. The packet delivery ratio of MANET sent traffic with normal network nodes is 1,900,000 bits/sec (1.9 Megabits) compare to misbehavior nodes the sent traffic is 1,800,000 bits/sec (1.8 Megabits). The received traffic MANET’s node is recorded first with normal network nodes 600,000 (75000 Kilobits) bits/sec compare with misbehavior nodes is 500,000 bits/sec (500 Kilobits). The retransmission retransmission attempt under misbehavior misbehavior nodes is lower than the normal network. This concludes the malicious nodes on network are not interested to retransmit retransmit the transmission transmission of network in order to participate participate in network throughput. throughput. The result showed the overall packets dropping occurred under misbehavior attack. OLSR routing table being dropped and not forwarded to other nodes in order to route maintain and route discovery phases. Simulation results for GRP under misbehavior nodes are examined. The throughput of the whole network under misbehavior nodes is 1,314,594 bits/sec (1.31 Megabits) where the original throughput of entire network is 3,400,000 bits/sec (3.4 Megabits) as it can observe the throughput of the network under misbehavior node has decreased dynamically and only 1.31 1.31 Megab Megabit itss is avail availabl ablee to whole whole netwo network. rk. Th Thee netwo network rk load load of the the netw network ork und under er misbe misbehav havior ior nod nodee 1,400 1,400,0 ,000 00 bits/ bits/sec sec (1.4 (1.4 Megab Megabit its) s) comp compare are to norma normall netw network ork load load 3,500,000 bits/sec (3.5 Megabits) is recorded, delay has increased during the misbehavior nodes as malicious node dropping the packet before forwarding it to other nodes on the network therefore any packet on network reach to its destination is dropped or reached late. The packet delivery ratio is compared against the normal network with misbehavior nodes, the the norma normall MANE MANET T traff traffic ic recei receive ved d 1,600 1,600,00 ,000 0 bits/ bits/sec sec (1.6 (1.6 Megab Megabit its), s), wher whereas eas with with misbehavior nodes (600 Kilobits), the MANET traffic sent with normal nodes is 1,450 bits/sec (1.45 Kilobits), compare to misbehavior node is 900 bits/sec (0.9 Kilobits). In GRP routing the total number of backtracks are higher with malicious nodes compare with normal nodes on the network. The GRP routing protocol is badly affected when malicious nodes does not pass the routing route to other node to network to find the route. The result of AODV routing protocol under misbehavior nodes examined, the packet delivery ratio of traffic sent is recorded as 15,000,000 bits/sec later introducing the misbehaivor nodes the MANET traffic sent is decreased to 13,686,074 bits/sec. looking at the traffic received received of normal normal node 2,800,0 2,800,000 00 bits/sec bits/sec with with misbeha misbehavior vior nodes nodes decrease decrease to 2,400,00 2,400,000 0 bits/sec bits/sec.. Similarly, the throughput of the network is decreased under misbehavior nodes, so network load as well.

Name: Sheraz Salim

Page 73


OLSR OLS R Routin Routing g Protoc Protocol ol Compar Compariso ison n with with Pulse Pulse Jammer Jammer and Misbeh Misbehavi avior or nodes nodes Attack Performance

Throughput

Net Networ work

Metric

bits/sec

bits/sec

loa load

Packet

Deliver Delay (sec)

Ratio Sent/Received bits/sec

1,500, 00,000 000 bi bits/sec sec

OLSR

1,3 1,300, 00,000

1,900,000 bits/sec

bits/sec

Normal

650,000 bits/sec

network

00,000 000 bi bits/sec sec puls pu lsee 1,200,

OLS OL SR

1,1 1,100, 00,000

1,800,000 bits/sec

bits/sec

Jammer

1,400,000 bits/sec

26 sec

450,000 bits/sec

attack OLSR

22 sec

1,200,000

1,827,631 bits/sec

Misbehavior

25.4 sec

550,000

nodes

GRP Routing Protocol Comparison with Pulse Jammer and Misbehavior nodes Attack Performance

Throughput

Network

Metric

bits/sec

bits/sec

load

Packet

Deliver Delay (sec)

Ratio Sent/Received bits/sec

GRP

Normal 3,400,000

bits/sec

Network GRP

3,500 3,500,0 ,000 00 bit bits/s s/sec ec

Pulse 1,000,000

1,300,000

Misbehavior

bits/sec

8.5 sec

1,600,000 bits/sec 1,000 1,000,0 ,000 00 bit bits/s s/sec ec

Jammer Attack bits/sec GRP

15,244 15,244,0 ,019 19 bit bits/s s/sec ec

1,900 1,900,0 ,000 00 bits/s bits/sec ec

29 sec

360,000 bits/sec 1,400 1,400,0 ,000 00 bit bits/s s/sec ec

1,800 1,800,0 ,000 00 bits/s bits/sec ec

23 sec

580,000 bits/sec

Nodes

AODV AODV Routin Routing g Protoc Protocol ol Compa Comparis rison on with with Pulse Pulse Jammer Jammer and and Misbe Misbehav havior ior nodes nodes Attack

Name: Sheraz Salim

Page 74


Performance

Throughput

Netw Networ ork k

Metric

bits/sec

bits/sec

load load Packet Deliver Ratio Delay (sec) Sent/Received bits/sec

AODV Normal Normal 5,000,000

bits/sec

Network AODV

Pulse 4,800,000

Jammer Attack bits/sec

2,000,000 bits/sec 2,200,000 bits/sec

AODV

5,500,000

2,200,000

Misbehavior

bits/sec

bits/sec

Nodes

15,000,000 bits/sec

15.3 sec

1,400,000 bits/sec 15,257,000 bits/sec

18.2 sec

1,400,000 bits/sec 13,686,047 bits/sec

14 sec

1,387,520 bits/sec

CHAPTER NO.6 6.1 Conclusion In this thesis the simulation study of our work consist of three routing protocol OLSR, AODV and GRP set up over MANET using medium FTP, medium E-Mail and low Database traffic analyzing their actions with respect to performance parameters, throughput, network

Name: Sheraz Salim

Page 75


load, and delay. The aim was to look at the performance under intelligent pulse jammer attack and under misbehavior nodes. Normal traffic is compared with different routing protocols, with pulse jammer attack and misbehavior nodes in terms of performance i.e. throughput, packet delivery ratio, delay, retransmission attempts. We showed several security breaches under pulse jammer attack and misbehavior nodes attack models using OPNET. Misbehavior nodes model attack showed the network affecting in several different security aspects, it not only degrading the performance of the entire network but also reflect the securit security y measure measuress like like availabi availability lity,, authent authenticat ication, ion, confiden confidential tiality ity,, integri integrity ty and nonrepudiation. Intelligent pulse jammer model showed the network degrading performance by generating noise on the radio frequency hence the jammer highlighted the security aspect and damaging damaging the throughput and made more complicated complicated for nodes to communicate on wireless wireless radio frequency. Several parameters affected in term of security under jammer attack.

6.2 Findings From this study we can conclude the protocols we chosen to implement on our network, is no single one having overall better performance under malicious activities i.e. Pulse Jammer attack and Misbehavior nodes. One Ad Hoc protocol may be better in term of routing whilst others may be better in term of throughput, packet delivery ratio, and delay. In our study we analysed Intelligent Intelligent Pulse Jammer attack and Misbehavior Misbehavior nodes attack with six different scenario along with performance metrics. We analysed the vulnerability of OLSR, GRP and AODV. Factor considered the performance of the routing protocols under Jammer and misbehavior attacks, OLSR has the worst performance compared with other two routing protocols AODV and GRP, in terms of throughput, network load and delay. One can almost be certain that AODV routing protocol performed better opposite to OLSR and GRP, its throughput, network load and delay is much better than the other routing protocols. AODV performed well under jammer attack and under misbehavior nodes attack. Based on the research we carry out and analysis analysis of the simulation results in OPNET Modeler Modeler 14.5, we describe the conclusion that OLSR and GRP is more vulnerable under pulse jammer attack and under misbehavior nodes attack.

Name: Sheraz Salim

Page 76


6.3 Future Work Significant research on MANET has done so far and are most likely to grow more and get huge presence in future communication infrastructure. As the use of MANET technology is increase security becomes a concern as a large number of research paper showed MANET are vulnerable when it come to security DoS attacks are launched against nodes easily. If we get more time to work on MANET security we shall implement cryptography to prevent MANET’s node from pulse jammer attack and from misbehaviour nodes attack for better and secure the security of MANET’s nodes. Other than that if we get a time of six to eight months we implement MANET security attack i.e. blackhole attack, wormhole attack, byzantine attack we shall also look for the mechanism to prevent from these attack.

Name: Sheraz Salim

Page 77


7. References 1 . H.

Bakht,

20 0 4 .

Understanding

Mobile

Ad

ho c

Network.

http://www.computingunplugged.com/issues/issue200406/00001301001.html [Accessed Feb – April 2010] 2. Imrich Imrich Chlamta Chlamtac, c, Marco Conti, Conti, Jennifer Jennifer J. N. Liu, Liu, 2003. Mobile Mobile ad hoc networking networking:: imperatives

and

challenges.

http://www.scribd.com/doc/19471793/MobileAdHocNetworkingImperativesandChall enges [Accessed Feb – April 2010] 3. B. Wu, J. Chen, J. Wu, M. M. Cardei, Cardei, 2006. A Survey of Attacks Attacks and Countermea Countermeasures sures in Mobile

Ad

Hoc

Networks.

http://student.f http://student.fau.edu/j au.edu/jchen8/web/ chen8/web/papers/S papers/SurveyBo urveyBookchapter. okchapter.pdf pdf [Accessed Feb – April 2010] 4. Ashikur Ashikur Rahman, Rahman, Pawel Pawel Gburzyns Gburzynski, ki, 2006. 2006. Hidden Hidden Problems Problems with the Hidden Hidden Node Problem.

http://citeseerx.ist.psu.edu/viewdoc/download?

doi=10.1.1.61.365&rep=rep1&type=pdf. [Accessed Feb – April 2010] 5. Yihong Yihong Zhou, Zhou, Dapeng Dapeng Wu, Scott Scott M. Nettles, Nettles, 2004. 2004. Analyzing Analyzing and Prevent Preventing ing MACMACLayer

Denial

of

Service

Attacks

fo r

Stock

802.11

Systems.

http://www.ar http://www.arias.ece. ias.ece.vt.edu/p vt.edu/publicati ublications/confere ons/conferences/UKC nces/UKC06.pdf 06.pdf [Accessed Feb – April 2010] 6. Cisco

systems

Networking

http://www.ci http://www.cisco.com/ sco.com/warp/pub warp/public/707/c lic/707/cisco-sa-20091 isco-sa-20091109-tls. 109-tls.pdf pdf [Accessed [Accessed Feb – April 2010] 7. Hasnaa Hasnaa Moustafa Moustafa and Houda Labiod, Labiod, 2005. 2005. Source Routin Routing-ba g-based sed Multica Multicast st Protocol Protocol for Mobile Ad hoc Networks. 8. H. Yang, Yang, H. Luo, Luo, F. Ye, S. Lu, and L. Zhang, Zhang, 2004. 2004. Secur Securit ity y in Mob Mobil ilee Ad hoc Networks: Challenges and Solutions, IEEE wireless Communication, Vol.11, Issue 1, pp . http://netlab18.cis.nctu.edu.tw/html/AdHoc_Network/slides/Chapter3-1-Security %20in%20Mobile%20Ad%20Hoc%20Networks.pdf [Accessed Feb – April 2010]

Name: Sheraz Salim

Page 78


9. Siddhar Siddhartha tha Gupte, Gupte, Mukesh Mukesh Singhal, Singhal, 2003. Secure Secure routing routing in mobile mobile wireless wireless ad hoc network, Deparment of computer science, university of Kentucky, Lexington, KY 40 50 8.

http://citeseerx.ist.psu.edu/viewdoc/download?

doi=10.1.1.117.1663&rep=rep1&type=pdf [Accessed Feb – April 2010] 10. Patroklos Patroklos G. Argyroudis and Donal O Mahohy, 2005. 2005. Secure Routing For Mobile Ad Hoc Hoc

Netw Networ orks ks,,

2005 2005..

http http:/ ://c /cit ites esee eerx rx.i .ist st.p .psu su.e .edu du/v /vie iewd wdoc oc/d /dow ownl nloa oad? d?

doi=10.1.1.81.4598&rep=rep1&type=pdf [Accessed Feb – April 2010] 11. Dagmara Dagmara Spiewak, Thomas Engel, and Volker Fusenig, 2007. Unmasking Unmasking Threats in Mobile

Wireless

Ad-Hoc

Network

Settings.

http://wiki.uni.lu/secan-lab/docs/Spiewak2007a.pdf [Accessed Feb – April 2010] 12. Houda Houda Labiod, Labiod, Mohamad Mohamad Badra, Badra, 2007. 2007. New technologie technologies, s, mobili mobility ty and security security,, springer. 13. Marianne Marianne Azer, Azer, Sherif Sherif El- Kassas, Kassas, Magdy El-Soud El-Soudani ani,, 2009. 2009. A Full Image Image of the wormhole attaks, towards Introducing Complex Wormhole Attacks in wireless ad hoc networks. 14. Rashid Rashid Hafeez Hafeez Khokhar Khokhar,, Md Asri Ngadi and Satria Satria Mandala, Mandala, 2008. A Review Review of Current

Routing

Attacks

in

Mobile

Ad

Hoc

networks.

http://eprints.utm.my/8213/1/NgadiMA2008_AReviewCurrentRoutingAttacksInMobi leAd-hocNetworks.pdf [Accessed Feb – April 2010] 15. A. Rajar Rajaram am,, Dr. Dr. S. Pala Palani niswa swami mi,, 201 2010. 0. Th Thee Trust Trust-B -Base ased d MACMAC-La Laye yerr Secu Securit rity y Protocol

for

Mobile

Ad

h oc

Networks.

http://www.enggjournals.com/ijcse/doc/IJCSE10-02-02-51.pdf [Accessed Feb – April 2010] 16. N. Shanthi, Shanthi, Dr. Lganesan Lganesan and Dr. K. Ramar, Ramar, 2010. Study of Differe Different nt attack on multicast mobile ad hoc network.

17. Erdal Cayirci, Cayirci, Chunming Rong, page 116, 2009. Book Security Security in wireless Ad Hoc and Sensor Network, John Wiley & Sons Ltd.

Name: Sheraz Salim

Page 79


18. Kama Kamansh nshis is and Md Md.. Liaka Liakat, t, 200 2007, 7, Secu Securit rity y Th Thre reat atss in Mo Mobil bilee Ad hoc Netw Network ork.. Master Thesis. Blekinge Institute of Technology. 19. 19. C. Siva Siva Ram Mu Murt rthy hy and and B. S. Mano Manoj, j, 2004 2004.. Ad Hoc Hoc Wire Wirele less ss Netw Networ ork, k, Architectures and Protocols, , chapter 9, page no 476. 20. Helena Helena Szczerb Szczerbicka icka,, Kishor Kishor S. Trivedi Trivedi and Pawan Pawan K. Choudha Choudhary, ry, 2003. Discrete Discrete Even Ev entt

Simu Simula lati tion on

with with Appl Applic icat atio ion n

to Comp Comput uter er

Comm Commun unic icat atio ion n

Syst System emss

Performances. 21. Jerry Jerry Banks, Banks, John S. Carson II, Barry L. Nelson, Nelson, David David M. Nicol, 2010. Discrete Discrete-Event System Simulation, Fifth Edition, Chapter No 1, page no 16. Pearson. 22. Rica Ricardo rdo F. Garz Garzia, ia, Mario Mario R. Garzi Garzia, a, 199 1990. 0. Netw Networ ork k Mo Mode deli ling, ng, Simu Simula lati tion, on, and Analysis, Chapter 1, page no 9, Dekker. 23. Lu Lucc Hogie Hogie,, Pasc Pascal al Bouvr Bouvry, y, Frede Frederic ric Guina Guinand, nd, 200 2006. 6. An Overv Overvie iew w of MANE MANETs Ts Simu Simula lati tion on,,

2006 2006..

http http:/ ://c /cit ites esee eerx rx.i .ist st.p .psu su.e .edu du/v /vie iewd wdoc oc/d /dow ownl nloa oad? d?

doi=10.1.1.106.3553&rep=rep1&type=pdf [Accessed Feb – April 2010] 24. Gunnar Gunnar Ronneber Ronneberg g and Olav Ly Lysne, sne, 2008. An OPNET OPNET-base -based d Simulat Simulation ion Mod Model el of SCISCI- nodes, nodes, Univers University ity of Oslo. Oslo. http:// http://cit citesee eseerx. rx.ist. ist.psu. psu.edu/ edu/view viewdoc/ doc/down download load?? doi=10.1.1.16.777&rep=rep1&type=pdf [Accessed Feb – April 2010] 25. S.A. S.A. Raza Razak, k, S. M. Furne Furnell, ll, P. J. Brook Brooke, e, 200 2003. 3. Atta Attacks cks again against st Mo Mobil bilee Ad Hoc Hoc Networks

Routing

Protocols,

university

of

Plymouth.

https://www.scm.tees.ac.uk/p.j.brooke/b/Razak+04a.pdf [Accessed Feb – April 2010] 26. Ola Mohamad, Rosilah Rosilah Hassan, Ahmed Patel, Patel, Rozilawati Razali, Razali, 2010. A Review of Security

Parameters

in

Mobile

Ad-Hoc

Networks.

http://www.ftsm.ukm.my/rosilah/files/OLA_ICICS.pdf [Accessed Feb – April 2010] 27. Th Thee Netw Network ork Simu Simula lator tor –NS2 –NS2,, discre discrete te event event simu simulat lator or targe targete ted d at netw network orkin ing. g. Internet : http://www.isi.edu/nsnam/ns/ns-build.html 28. OPNET

Technologies,

Inc.

“OPNET

Modeler”,

Internet

:

http://www.opnet.com/solutions/network_rd/modeler.html 29. GloMoSim GloMoSim discrete simulation simulation environment for wireless and wired network systems, systems, Internet http://pcl.cs.ucla.edu/projects/glomosim/

Name: Sheraz Salim

Page 80


30. QualNet QualNet Network Simulator, Simulator, Internet : http://www.sc http://www.scalable-net alable-networks.co works.com/ m/ 31. 31. OMNe OMNeT+ T++ + open open-s -sou ourc rce, e, comp compon onen entt base based d simu simula lati tion on pack packag agee for for mode modeli ling ng communication network. Internet http://www.omnetpp.org/ 32. Mishr Mishraa Amit Amitabh abh,, Nadka Nadkarni rni Ketan Ketan M., M., and Ilya Ilyass Mo Moha hamm mmad, ad, 2003.“Chapter 2003.“Chapter 30: Security in wireless ad-hoc networks, the handbook of Ad hoc wireless network”. , CRC PRESS Publisher. 33. Uyen Uyen Trang Trang Nguyen Nguyen and Xing Xiong, Xiong, “Rate adaptive adaptive Multicast Multicast in Mobile Mobile Ad hoc Netwo Networks” rks”,, IEEE IEEE Internat Internationa ionall Conferen Conference ce on Ad hoc and Mobile Mobile Comput Computing, ing, Networking and Communications, WiMob, Montreal, Canada, 2005 34. Sajjad Ali & Asad Ali, Performance Analysis of AODV, AODV, DSR and OLSR OLSR in MANET, Department of Electrical Engineering with emphasis on Telecommunication Blekinge Institute

of

Technology,

Sweden

20 09 .

http://www.bth.se/fou/cuppsats.nsf/all/252aefb4936b9db3c12576b20053b8a5/$file/Pe rformance%20Analysis%20of%20AODV,%20DSR%20and%20OLSR%20in %20MANET.pdf [Accessed July – August 2010]

Name: Sheraz Salim

Page 81


OPNET_MODELER_MANET Attacks Thesis

Recommend Documents