Application of comprehensive learning particle swarm optimisation algorithm for maximum likelihood DOA estimation in wireless sensor networks

Abstract

Direction of arrival (DOA) estimation is one of the challenging problem in wireless sensor networks. Several methods based on maximum likelihood (ML) criteria have been established in literature. But, the multimodal nature of ML cost function is one of the inherent limitations in ML-DOA estimation technique. Generally, to obtain the likelihood solutions, the DOAs must be estimated by optimising a complicated function over a high-dimensional problem space. Recently particle swarm optimisation (PSO) algorithm is used in MLDOA estimation. To overcome the drawback of premature convergence in PSO, a learning strategy is introduced, and this approach called comprehensive learning particle swarm optimisation (CLPSO), which is applied to this problem and a comparison of results is made between these two. Simulation results confirms that the ML-CLPSO estimator is significantly giving better performance at low signal-to-noise ratio compared to conventional methods like multiple signal classification (MUSIC) and ML-PSO in various scenarios at less computational costs.