SHAPARS: Shadowing past reputation system based on subjective logic for mobile ad hoc networks

Abstract

Cooperation of nodes in forwarding packets for each other is essential in setup and operation of mobile ad hoc networks. Selfish behavior of nodes in such networks may degrade overall performance and lifetime of the network dramatically. Therefore, detection and prevention of selfish behavior is of great importance in this field. In this paper we propose SHAPARS, an efficient subjective logic based reputation system to detect and isolate selfish nodes in mobile ad hoc networks. It prevents selfish mobile nodes from hiding their selfishness by moving around in the network. It also uses an enhanced definition of expectation in subjective logic to speedup selfish node detection. Simulation results show considerable improvements in terms of decreasing success rate of selfish nodes and total number of reputation queries, compared to previous related works.