DEVELOPMENT OF AN ARTIFICIAL FISH SWARM ALGORITHM BASED ENERGY EFFICIENT TARGET TRACKING SCHEME IN WIRELESS SENSOR NETWORKS

By

Author

Presented To

Department of Engineering

ABSTRACT
Optimal deployment of sensor node in other to ensure optimum network coverage is one of the challenging problems faced by Wireless Sensor Network (WSN) researchers due to the complexity and exhaustive nature of WSN. Target tracking problem is concerned with maximizing the lifetime of the network while continuously monitoring a set of targets.This dissertation presents an optimal deployment of WSN and target tracking scheme using the intelligent swarming behaviours of Artificial Fish Swarm Algorithm (AFSA). The preying, swarming and chasing behaviours of AFSA were initially replicated using MATLAB R2013b simulation environment. The position of network nodes were randomly deployed in a network coverage area of 60 square meters with a total of 60 sensor nodes of 4m radius and communication range of 15m using the replicated AFSA algorithm. Thereafter, the replicated AFSA was used to detect event based on target discovery probability model. A series of simulation were performed, and results showed that the proposed technique can attain maximum network coverage of 77.87% when the number of iteration was 25 after which it kept an almost constant value for the rest of the simulation process. The relationship between network coverage and number of mobile nodes also showed that network coverage increased with increase in mobile nodes. The approach indicated maximum network coverage of 80.07% when the mobile node was 50. Thereafter, it tended towards stability when the number of network nodes was above 50. Effects of various attenuation factors on the proposed model were evaluated and simulation results shows that the proposed method successfully attains maximum network coverage of 70.58%, 70.99%, 72.69% and 77.15% when the attenuation factors are 0.75, 0.8, 0.85 and 0.90 respectively. Target tracking simulation scenarios were presented and results showed that the computation energy required to successfully track 30, 45 and 60 targets were 21.63%, 28.003% and 36.99% less than the energy (time taking) required to track the 15 targets respectively.