Enhancing Security with Effective Routing Mechanisms in Ad Hoc Networks

Abstract

Ad hoc networks are special class of networks which are autonomous in nture. Coor- dination of communication in ad hoc networks is inherently distributed. Nodes (com- municating devices) deploy automatically as they configure themselves. The nodes function cooperatively to perform self deployment in absence of any central entity. Hence, ad hoc networks are used for specific applications, like disaster management, location based services, tactical areas etc.; where prior deployment of any network is not foreseen. This endangers the ad hoc network since it is exposed to several security threats. Since no centralized infrastructure is available for coordination, nodes take decisions on their own and hence, there is severe threat to the performance of the network. As malicious nodes hamper performance of network; tactical applications need to detect the malicious nodes during real-time communication. Most of the work done till date, focus on detecting and specifying the presence of malicious nodes in the network. A node is said to be malicious, if it is not forwarding packets or control messages, ooding the network with unnecessary route requests or data packets, and not cooperating in the network for network related activities. The work is focused on, how well is the Intrusion Detection or Prevention System accurate in detecting the presence of malicious nodes. Hence, optimizing the routing in presence of attacking nodes is difficult. Moreover, none of the simulators are holding required plug-ins to create various attacks along with the specification that suits requirement of the tactical applications. Primary focus of the research work is to create various attacks like blackhole at- tack, ooding attack, dropping routing trafic attack, selfshness attacks and anoma- lous attacks. These attacks have been simulated and analyzed. The next most im- portant aspect of performance of an ad hoc network is that routing should not suffer even if the malicious nodes are present in the network. Only then, it is possible to perform the e ective routing. The major contribution is in identifying the suitable methodology; using which, the malicious nodes are identi ed and simultaneously they are mitigated during the routing process. Timed Finite State Machine (TFSM) based Intrusion Detection Systems (IDS) are implemented and incorporated with the stan- dard Ad hoc On-demand Distance Vector (AODV) routing protocol to deal with such intrusions simultaneously. Once the network is deployed and the malicious nodes are present in the network, it is important to discuss about how much the nodes trust each other and how can the nodes be forced to cooperate in the network. For this, a routing technique is proposed and implemented, which is trust based routing mechanism. This trust develops on the basis of how the nodes are behaving in general during the temporal communication among nodes in the network. In a nutshell, the contribution of the work is a solution to identify availability attacks and handle them during the routing process; so that the gravity of such attacks get mitigated upto a certain extent.