SOPC (system on the programmable chip) based array signal processing algorithm

The traditional algorithm of DOA (Direction of Arrival) angle is based on MUSIC (Multiple Signal Classification) algorithm. Such an algorithm, as is well known, has poor performance in the case of multiple interrelated signals, low signal noise ratio (SNR) or several objects whose azimuth angles are near. And the computation of MUSIC algorithm always needs massive processing time. How to improve the performance of the MUSIC algorithm and design a suitable hardware platform is extremely difficult. We used an improved algorithm of MUSIC to overcome this difficulty. Several difficult situations are discussed. First, when signal sources are interrelated, subsection 1.2 of the full paper explains how our improved MUSIC algorithm could find the DOA of each of several objects. Second, when the SNR is low, subsection 1.2 explains that the improved MUSIC algorithm could also find out the DOA of each of several signal sources. Third, even if the signal sources have azimuth angles that are near, still subsection 1.2 explains that the improved MUSIC algorithm could find out the DOA. And subsection 2.1 explains that we have developed and tested a hardware platform for the improved MUSIC algorithm. It employs SOPC and FPGA (Field Programmable Gate Array) technologies to satisfy the requirements for more advanced performance. Having both the embedded soft kernel processors and hardware acceleration units in the platform enables fast processing of massive signals.