Markovian process based reputation mechanisms for detecting selfish nodes in MANETs: A survey

Abstract

Reputation is considered to be an important aspect for enabling reliable data communication among the mobile nodes present in a multihop network like mobile ad hoc network. But, the presence of the selfish nodes could wreak considerable damage to the network performance. Reputation based models were mainly formulated for enforcing cooperation and at the same time for punishing misbehaving nodes like selfish nodes. At the same time, vast numbers of mechanisms proposed for detecting selfish nodes were greatly based on the consistent history of interactions between the mobile nodes existing in an ad hoc environment. In this paper, we survey and classify notable works contributed for detecting selfish nodes based on Markovian processes. These Markov based mechanisms detects the selfish nodes based on the present behavior of nodes determined in discrete or continuous time rather than considering their past history. Finally, we discuss on the merits and demerits of the existing Markovian models available in the literature.