Research On Algorithm Of Coherent Signals Based On Uniform Circular Array

Coherent interference is a common problem in the signal processing. The presence of coherent signals will lead to performance degradation, even invalidation of the DOA(Direction of Arrival) estimation algorithm. Hence, the DOA estimation of the coherent signals is an important aspect of the array signal processing. Most DOA algorithms for coherent signals are proposed based on the special array manifold of uniform linear array(ULA). However, the line array has small range of direction finding and different estimation accuracy in different directions. Uniform circular array(UCA) which provides direction finding of 360 °can overcome the defects of small range for direction finding of uniform linear array, at the same time the estimation accuracy is approximately the same in each direction. Therefore, this research studies the DOA estimate algorithm of coherent signals based on UCA, the main contents are summarized as follows:Firstly, the conclusion that the coherent signals will lead to the MUSIC algorithm failure is verified through theoretical analysis and computer simulations. We focus on the DOA estimation algorithms for coherent signals based on the MUSIC algorithm including spatial smoothing algorithm, eigenvector reconstruction algorithm and data vector reconstruction algorithm. We compared the estimation performance of aforementioned algorithms based on the ULA. Then an improved data vector reconstruction algorithm is proposed which solves the problem existing in the original one that the reconstructed data matrix is not hermitian matrix. The simulation results show that the performance of the improved data vector reconstruction algorithm is better than the original algorithm.Secondly,the DOA estimation algorithms for solving the coherent signals require mode transition when they are used the UCA. We compared the estimation performances of the spatial smoothing algorithm, the eigenvector algorithm, the data reconstruction algorithm and the improved data reconstruction algorithm and studied the estimation performances under different phase differences of the incident signals. The simulation results show that the performance of the improved data vector reconstruction algorithm is better than the original algorithm. We also simulate the effects of mode number on the performance of the algorithms, the simulation results show that appropriate mode number can increase the performance of the algorithms.