A fast and hierarchical source localization algorithm for planar spiral array implemented using GPU

Accurate and fast localization of multiple sources is an important issue in many applications such as fault diagnosis. The well-known steered response power (SRP) method is widely used in the source localization but exhibits high computational expense. Therefore, this paper investigates a coarse-to-fine steered response power source localization algorithm to speed up the localization process. The detailed comparisons with previous algorithms are made to demonstrate that the proposed scheme is much faster, robust, and accurate. In addition, the algorithm is implemented in real time based on CUDA frame (Compute Unified Device Architecture) using GPU with high-parallel execution.