Abstract

Localization is one of the important technical issues of wireless sensor networks. Based on the traditional genetic algorithm applied in node localization, this paper introduces pattern search algorithm for the selection of population, meanwhile, genetic operators are designed. The theoretical analysis and simulation results show that, even though the proportion of anchor nodes and network connectivity is poor, and distance measurements are corrupted by Gaussian noise, the improved algorithm achieves high accuracy. Thus, it can be applied to most large-scale wireless sensor networks for node localization.

Paper Details