Optimized placement of nodes for target detection in sensor networks

Abstract

In order to improve the precision of the target detection in wireless sensor networks, a new approach based on genetic algorithm (GA) was proposed to optimize the placement of the sensor. The target location problem was transformed into locating a target at a grid point through modeling the sensor field as a grid of points. Moreover, the sensor placement problem was formulated as a combinatorial optimization problem, which is aimed at minimizing the maximum discrimination error under the restraints of limited cost and complete coverage. The GA approach uses binary coding to represent the location, and both single parent cross-over operator and single parent mutation operator are used to improve its speed and efficiency. Experimental results have shown that a global optimal solution can be quickly obtained using the proposed method and the precision requirement for target location is satisfied.