Accelerating MUSIC method on reconfigurable hardware for source localisation

Abstract

This paper presents an implementation of the high resolution source localisation method MUSIC on an FPGA system. This method exploits the eigenvalue decomposition (EVD) of the correlation matrix generated from the signals received at different sensors. An efficient architecture for the computation of the singular value decomposition (SVD) and the EVD based on the Brent, Luk, Van loan (BLV) systolic array has been proposed. The architecture is three times more efficient and faster than the existing BLV structure. An optimised implementation has been efficiently carried out on the PPRC1000 board using a high level language for hardware design "Handel-C".