Novel implementation technique for a wavelet-based broadband signal detection system

摘要

This thesis reports on the design, simulation and implementation of a novel Implementation for a Wavelet-based Broadband Signal Detection System. There is a strong interest in methods of increasing the resolution of sonar systems for the detection of targets at sea. A novel implementation of a wideband active sonar signal detection system is proposed in this project. In the system the Continuous Wavelet Transform is used for target motion estimation and an Adaptive-Network-based Fuzzy inference System (ANFIS) is adopted to minimize the noise effect on target detection. A local optimum search algorithm is introduced in this project to reduce the computation load of the Continuous Wavelet Transform and make it suitable for practical applications. The proposed system is realized on a Xilinx University Program Virtex-II Pro Development System which contains a Virtex II pro XC2VP30 FPGA chip with 2 powerPC 405 cores. Testing for single target detection and multiple target detection shows the proposed system is able to accurately locate targets under reverberation-limited underwater environment with a Signal-Noise-Ratio of up to -30db, with location error less than 10 meters and velocity estimation error less than 1 knot. In the proposed system the combination of CWT and local optimum search algorithm significantly saves the computation time for CWT and make it more practical to real applications. Also the implementation of ANFIS on the FPGA board indicates in the future a real-time ANFIS operation with VLSI implementation would be possible.