Enhanced Grade Diffusion for Node Recovery in Wireless Sensor Networks

Abstract

Wireless Sensor Networks (WSNs) are the most widely used networks to monitor distributed remote environments. In WSNs, when nodes lose touch within the network, they fail to operate. There are various reasons due to which this can occur, such as, in the event of software crash or hardware failure, the failures of a communication system or the loss of network connectivity occur. Any failure or disruption or split of network region that stops the node from communicating with the network is considered as node failure. Number of researchers has addressed the issue of recovering the Faulty Nodes (FNs). The aim of the Grade Diffusion (GD) algorithm is to recover the FN by rerouting the whole path. This gives quick solution to the problem but is very expensive in terms of power efficiency of the whole network as the rerouted path can be comparatively long and requires more energy to route packets. Recovery of nodes using different mechanisms like LeDiR (Least-Disruptive topology Repair) and GD has been the main solution to the node failures in the running networks. In this thesis GD based Particle Swarm Optimization (PSO) is implemented to re-route the paths. That originally reroutes the path by selecting the best path through consideration of the situational best replacement of the node in place of failing node. Such path selection often leads to bad distribution of energy in the network in later stages. Usage of PSO allows energy distribution in the network uniformly to utilize the maximum lifetime of the Wireless Network. The optimized path selection using PSO allows the protocol to maintain energy levels in the network by choosing the path which is most efficient in terms of energy consumption. The comparative analysis of the proposed technique has been done with the existing techniques. The results have been evaluated in the terms namely, network lifetime, delay, throughput and energy. The implementation of routing protocol, node failure and optimization of paths using GD has been simulated in MATLAB