An Adaptive and Predictive Security Model for Mobile Ad hoc Networks

Abstract

Mobile ad hoc networks are infrastructure-free, pervasive and ubiquitous in nature, without any centralized authority. These unique characteristics coupled with the growing concerns for security attacks demand an immediate solution for securing the ad hoc network, prior to its full-fledged deployment in commercial and military applications. So far, most of the research in mobile ad hoc networks has been primarily focused on routing and mobility aspects rather than securing the ad hoc networks themselves. Due to ever increasing security threats, there is a need to develop schemes, algorithms, and protocols for a secured ad hoc network infrastructure. To realize this objective, we have proposed a practical and effective security model for mobile ad hoc networks. The proposed predictive security model is designed using a fuzzy feedback control approach. The model is based on identifying critical network parameters that are affected by various types of attacks and it continuously monitors those parameters. Once we measure the relative change in these parameter values, we could detect the type of attack accurately and protect the system, without compromising its effectiveness. Experimental results of the model simulated for selected packet mistreatment attacks and routing attacks are very promising.