Implementation Of Direction Of Arrival (DOA) Estimation System Based On Multiple Signal Classification (MUSIC) Algorithm

The MUltiple SIgnal Classification MUSIC algorithm is a kind of DOA (Direction Of Arrival)estimation technique based on eigenvalue decomposition, which is also called subspace-basedmethod. This method is well known as high resolution DOA estimation technique. In the case ofmultipath fading channel, the performance of MUSIC decrease dramatically, which is alwaysbeen existed; so other techniques should be added to enhance MUSIC capability.Forward/Backward Spatial Smoothing (FBSS) technique is one of the most popular methods usedin literature to enhance MUSIC performances in such undesired situations. In addition of its superresolution capability, MUSIC is very suitable for integration on logic circuit devices such asFPGAs (Field Programmable Gate Array) technologies.According to the situations above, the work of the thesis is as follows. First, Basic principlesof DOA estimation based on MUSIC algorithm are introduced including channel propagationcharacteristics and mathematical formulation for DOA estimation based on MUSIC. Simulation isthe transitional stage between theory and mathematical derivations on the one hand, and theimplementation of a real system on the other. On these bases, a simulation platform is designedand the performances of MUSIC are evaluated for many simulation Scenarios for both of Basicalgorithm and MUSIC with FBSS. Simulation results are compared with other theoreticalapproximations and theoretical limits such as Cramer-Rao Bound (CRB).Further, a hardware design of Unitary MUSIC algorithm has been developed in this thesis,and its Xilinx System Generator (XSG) based hardware implementation was also introduced.This Unitary algorithm uses CORDIC (COordinate Rotation DIgital Computer)-based TriangularSystolic Array for QR-decomposition to deal with EVD (eigenvalue decomposition). TheMUSIC spectrum is computed with spatial DFT (Discrete Fourier Transform) using FFT blockoffered by Simulink-Xilinx blockset library. Comparing with MATLAB based system’s results;the proposed design has a super resolution capability.