Research Of DOA Estimation Algorithms For Wideband Signals

With the increasing of the signal complexity, the wideband signal processing has become a very important research direction. Wideband array signal processing is an important branch of signal processing. Direction of Arrival (DOA) estimation of wideband signals is widely used in communication?radar?sonar and other military and civilian fields.In the thesis, by using Uniform Circular Array (UCA) and Uniform Linear Array (ULA) as the receiving array, DOA estimation of wideband signals are investigated in-depth. The main work is as follows:First, the mathematical models of narrowband array signal, wideband array signal and correlated array signal are established, and two classical subspace-based DOA estimation algorithms for narrowband array signal,MUSIC and ESPRIT, are investigated. The key idea of MUSIC algorithm is using the orthogonality between noise subspace and signal subspace to estimate the angle of incidence. ESPRIT algorithm uses the rotation invariance of two identical sub-matrixes to estimate the angle of incidence. The performance of the two algorithms is evaluated by computer simulation.Second, the DOA estimation scheme for wide-band sources is designed, and the associated algorithms for focus conversion and the estimation of the number of unknown source are studied. The structure basis and method of TCT and RSS focus matrices are introduced and investigated. Gerschgorin algorithm and the information theory algorithm are used to estimate the number of sources in space, which is needed for subspace decomposition. The performance and feasibility of algorithms is proved by simulation results.Finally, since the estimation error of the source number will affect the performance of DOA estimation, a wide-band DOA estimation algorithm without source number is proposed in the thesis to solve this problem. The proposed algorithm uses beam-forming theory to get the weight vector, and uses the output of array to estimate DOA. The simulation results shown that the proposed algorithm has good performance and is robust to the variation of signal to noise ratio.