Modeling and Detection of Flooding based Denial of Service Attacks in Wireless Ad Hoc Networks using Uncertain Reasoning
ABSTARCT :
Wireless Ad Hoc Networks are self-organizing networks deployed without any centralized infrastructure. Flooding based Denial-of-Service (DoS) attacks were targeting the constrained resources of mobile nodes as well as the network. Route Request (RREQ) flooding attack is one of the prominent DoS attacks launched from the network layer in which the attacker sends a huge number of spoofed Route Request (RREQ) packets which not only overflows the target buffer but also creates network congestion. In this work, novel methods were suggested for defending RREQ flooding attack in Wireless Ad Hoc Network using two well-known frameworks in uncertain reasoning namely Bayesian Inference and Dempster-Shafer (D-S) evidence theory. The present work reports the modeling of RREQ traffic and developed an optimum algorithm for the detection of persistent RREQ flooding attack using Bayesian Inference. The algorithm was further refined for the detection of high rate and low rate pulsed RREQ flooding attack using D-S evidence theory. Based on the comprehensive evaluation using mathematical modeling and simulation, the proposed method successfully defended any type of flooding based DoS attack in Wireless Ad Hoc Network with lower communication and memory overhead.
EXISTING SYSTEM :
? Network forensics is still under active investigation by the research community, especially to address the issues in wireless networks.
? Mobile Ad hoc network (MANET) a kind of wireless networks.
? It is the distributed systems having wireless mobile nodes that can freely and dynamically selforganise into arbitrary, temporary, and ad hoc network topologies, allowing connections within the network neither having pre-existing communication infrastructure nor centralized administered control management.
? If no route from source node to destination node exists in source route cache, then source node initiates Route Discovery and sends out a ROUTE REQUEST message to find a route.
DISADVANTAGE :
? All kinds of networks, from traditional infrastructure networks to wireless self-organized networks, have a variety of security issues crossing all layers in the network architecture.
? Therefore, the critical research problem is how to assess each node in the networks precisely with dynamic evidence in a distributed manner.
? Uncertain reasoning was initially proposed from the artificial intelligence community to solve the problems in expert systems, which have frequent counter-factual results.
? Cognitive radio (CR) has become a promising technology to deal with the spectrum shortage problem.
PROPOSED SYSTEM :
• In the proposed work, forensic analysis is done using fuzzy logic. Motivation of using fuzzy logic is that, through fuzzy logic more appropriate pattern analysis rules can be implemented for both DoS and DDoS due to RREQ flooding.
• In MANET each node are self capable that support routing functionality in an ad hoc scenario, forwarding of data or exchange of topology information using wireless communications.
• Flooding attack causes excessive volume of traffic to deplete key resources of the target legitimate users, since the system get congested so forth, there is denial of services.
• The study of network forensics analysis for attacks in wireless network and in MANET is considered still in progress.
ADVANTAGE :
? These inside attacks always emerge in all kinds of networks and have significant impact on the performance of networks.
? Although the number of packets received correctly decreases as long as the number of nodes increases, the performance of our scheme has a big improvement.
? Through simulations, we show that the proposed scheme has a better performance in trust evaluation comparing to existing schemes.
? However, the proposed scheme has a better performance than the existing one because the trust values of malicious nodes in the proposed scheme are more accurate than its counterpart.
? The throughput performance of MANETs with the consideration of a variety of causes for packets dropping can be improved and the false alarm probability of the proposed scheme is lower.
|