Distributed binary consensus algorithm in wireless sensor networks with faulty nodes

Abstract

In sensor networks, consensus is a procedure to enhance the local measurements of the sensors with those of the surrounding nodes, and leads to a final agreement about a common value. The question here is how we can achieve the the consensus in a large network containing some faulty nodes. In this paper, we present distributed binary consensus algorithm (BCA) over the wireless sensor networks (WSN) in presence of faulty nodes. With binary consensus, each sensor node, observes one of two states TRUE and FALSE and the aim is to decide which one of the two states was held by the majority of the nodes. We details the implementation of the distributed BCA in WSN when the network contains P faulty nodes. The implementation was tested on sensor nodes using the TinyOS Simulator (TOSSIM) for a WSN with a large number of nodes. Here, TOSSIM guarantees that the code performs correctly when deployed on the physical nodes. In performance evaluation, we consider the analysis of the average convergence time over the simulated environment and considering the presence of P malicious nodes. These results are presented for a WSN with different topologies such as fully connected, path, ring, Erdos Reny random, and star-shaped.