Fault Node Recovery Algorithm for a Wireless Sensor Network

Abstract: This project proposes a Distributed Dynamic Diffusion Based Algorithm [DDBA] to enhance the life time of a wireless sensor network [WSN] when some of the sensor nodes stop their working. This proposed algorithm is based on concurrent working of genetic algorithm [GA] and grade diffusion [GD] algorithm. This algorithm will increase the number of active nodes by reducing the rate of data loss and rate of energy consumption. The grade diffusion algorithm will generate the neighbor nodes, grade values, routing table and payload value for each sensor node. The genetic algorithm will search for the non-functioning sensor nodes and it will replace the non-functioning sensor nodes by selected functional nodes in the network. The algorithm is based on the battery levels, location and load on the node, and its goal is to divide the power consumption within the network so that the majority of the nodes consume their power supply at relatively the same rate regardless of their exact location. This leads to maximization in terms of system maintenance, such as replacement of the batteries all at once than doing it one by one, and maximizing the overall system performance by allowing the network to function at 100% capacity throughout most of its lifetime instead of having a steadily decreasing node population. The system distributes the lifetime and load on the network better than the previous approach.

Keywords: Genetic Algorithm, Grade Diffusion (GD) Algorithm, Wireless Sensor Networks (WSN).