Consensus in sensor networks in presence of hybrid faults

Consensus in the presence of faults is a challenging task, particularly for wireless sensors that are prone to failures and are often deployed in inhospitable terrains. Consensus in Wireless Sensor Networks (WSNs) is possible if the link-densities among the neighboring nodes are adequate such that the effect of failures due to various types of faults can be masked. The theoretical results in literature place strict requirements on the link-density of each sensor (node) as the prerequisite to consensus. Furthermore, these results are not scalable in large randomly distributed WSNs. This research investigates the consensus problem, herein referred to as Global Convergence (GC), to determine the effect of varying link-densities on GC in the presence of faults and to design algorithms that are better suited to reach GC. The intertwined relationship among various parameters that affect GC are investigated as well. To achieve these goals, a GUI simulator has been developed that illustrates the effectiveness of the algorithms designed and that provides better insight into the application of GC to real-world applications in large WSNs.