A Global Reputation Estimation and Analysis Technique for detection of malicious nodes in a Post-Disaster Communication environment

Collection and analysis of situational data in a post disaster scenario is crucial for providing effective relief operation in the disaster stricken areas. However, malicious and selfish behavior of entities that forward such data pose to be a serious threat against transmission of sensitive situational data in a Post Disaster Communication Network. Due to the highly distributed nature of such a network and absence of trusted third party, one has to depend on attributes like trust and reputation of a node for evaluating it as honest and altruistic. However, a node cannot be expected to have knowledge about the global reputation of all other nodes in a distributed network. For this, we propose a scheme called GREAT (Global Reputation Estimation and Analysis Technique) that uses statistical estimation technique to estimate the global reputation of a node as a forwarder and as a rater from sample reputation values collected from a sample set of nodes in the network. GREAT eventually identifies selfish and malicious nodes in the network and excludes them to a great extent from future communication activities.