Bayesian-based Trust Initialization for Reputation Management in Wireless Sensor Networks
Small Business Information
15400 Calhoun Drive, suite 400, Rockville, MD, -
AbstractABSTRACT: Trust and reputation management system has been proven to be an effective method to solve many security issues in Wireless Sensor Networks (WSN). A typical reputation management system evaluates the trustworthy of a sensor node based on historical interactions. However, trust initialization remains to be a challenging issue, as there are no historical interactions in the initialization phase of the reputation management system. In this effort, we propose to develop an effective Trust Initialization Mechanism (T-INIT) based on Bayesian Fusion for evaluating the initial trustworthy of WSNs. The objective of this effort is to improve existing trust initialization strategy by taking a set of context parameters into consideration. Our final target is, based on our studies in Phase I, to build an advanced trust initialization strategy integrating with existing reputation/trust management system, so that we can successfully minimize the warm-up period and achieve the full advantages of the trust/reputation. BENEFIT: The proposed T-INIT system provides a solid solution for robust, efficient and accurate trust initialization for wireless sensor networks. We expect to produce a software prototype equipped with distributed reputation management system for military as well as researchers in the networking community to investigate reputation management in dynamic wireless networks. First, the proposed T-INIT solution has tremendous applications potential in dynamic military applications. Given the GIG vision, such heterogeneous and dynamic wireless sensor webs will be common and therefore secure, robust, efficient and timely routing solution is necessary. The proposed architecture, algorithms, and the developed software tool can be applied to various military networks for major programs like layered sensing program, surveillance network, JUMPS, DCGS, etc. Second, due to the increasing popularity of ubiquitous computing technologies, our proposed solution can be applied into a large number of commercial network applications, such as industrial control networks, disaster networks, and border monitoring networks. The size of the market is quite large and may grow rapidly with the commercial demand in network reliability and availability.
* information listed above is at the time of submission.