Load Balancing in Wireless Sensor Network using Divisible Load Theory

Abstract

In this thesis, optimal load allocation strategies are proposed for a wireless sensor network which is connected in a star topology. The load considered here is of arbitrarily divisible kind, such that each fraction of the job can be distributed and assigned to any processor for computation purpose. Divisible Load Theory emphasizes on how to partition the load among a number of processors and links, such that the load is distributed optimally. Its objective is to partition the load in such a way so that the load can be distributed and processed in the shortest possible time. The existing strategies for both star and bus topologies are investigated. The performance of the suggested strategy is compared with the existing ones and it is found that it reduces the overall communication and processing time if allocation time is considered in the previous strategies.