Low complexity subspace fitting method for wideband signal location

Abstract

Weighted subspace fitting (WSF) method has superior performance compared to other subspace-based techniques in broadband signal location, but it is a heavy computational task. An adaptive Particle Swarm Optimization algorithm (PSO) is proposed to solve nonlinear global optimization problem of WSF. The multi-stage Wiener filter is applied to construct the aim function for complexity reduction. Since DOA is estimated without eigen-decomposition and multidimensional search is done in an intelligent way, significant reduction in complexity is achieved. Simulation results demonstrate the effectiveness of the proposed methods.