Evaluating MANET Uncertainty Analysis Framework

Abstract

Using trust relationships between nodes, decision making in a peer-to-peer distributed network like MANET possible to improve. Uncertainty plays an important role in such decision making process. Our earlier research work viz. Sandeep et al. [1] proposed Uncertainty Analysis Framework (UAF) for MANET and integrated it with different trust variants representing direct trust, indirect trust and global trust. The research work explored relationship between uncertainty and trust. The UAF is found useful to quantify Belief, Disbelief, and Uncertainty (BDU) of a network. The earlier research work measured impact of direct, indirect and global trust on the performance of the routing protocols. The proposed work is useful to get more insights about - MANET working, effectiveness of trust based algorithms and impact of uncertainty on the behavior of MANET. This research work analyses UAF behavior and performance of indirect and global trust-based routing protocols using different test conditions. The proposed research work observed that, uncertainty reduces drastically when trust is used for decision making in routing protocols. It is also observed approximately 140 seconds are required to identify good and misbehaving nodes in the network. Belief, Disbelief and Uncertainty values are fairly stable after 140 seconds from network bootstrap. Indirect trust model is found more useful for open and distributed peer-to-peer networks.