A Study Of Narrowband Coherent Signals DOA Estimation Algorithm

Direction of arrival(DOA) estimation is one of the most important research directions in array signal processing, which is widely used in mobile communications, sonar, seismology, radar and other fields. At present, Scholars at home and abroad have proposed many classic super-resolution DOA estimation algorithms. However, in the real environment, due to multipath propagation and artificial disturbance, the incoming signals include coherent signals, and the classic DOA estimation algorithms are incapable of estimating the DOA of coherent signals effectively. Aimming at DOA estimation of coherent sources, the researchers also proposed a lot of decoherence algorithms, but most of them can not meet the requirements of practical application background. Higher precision, smaller amount of calculation, higher utilization of sensors and better robustness are the developing direction of coherent sources DOA algorithm, this paper proposes several kinds of more effective improved algorithms base on the current common decoherence algorithms.This thesis focuses on DOA estimation of coherent narrowband signals. First of all, the mathematical models of coherent signals and independent signals are established, some traditional DOA estimation algorithms are discussed, including Conventional Beam-forming(CBF) method, the minimum variance(MVM) method, MUSIC algorithm, ESPRIT algorithm, at the same time the principles and characteristics of the traditional DOA estimation algorithm are analyzed. Computer simulation results show that, apart from beam-forming method, the other algorithms are unable to estimate the DOA of coherent signals. Secondly, the principles of current common decoherence DOA estimation algorithms are studied, including space smoothing algorithm, front-rear space smooth algorithm, feature vector matrix reconstruction algorithm, the cross-correlation vector matrix reconstruction algorithm, Toeplitz matrix reconstruction algorithm, modified MUSIC algorithms. The space smoothing algorithm and the vector matrix reconstruction algorithm are of high accuracy, but the array aperture utilization rate of them is low. The toeplitz matrix reconstruction algorithm is a high robustness method, but its estimation accuracy is bad. The modified MUSIC algorithm is able to decoherence, but the number of signal source which can be estimated is limited. Finally, three improved decoherence algorithms are proposed based on classical algorithms:1. Taking advantage of the Toeplitz matrix decoherence theory, an improved matrix reconstruction algorithm is proposed, which is of high senor utilization rate and performs better than space smoothing algorithm for low SNR.2. Combined with spatial smoothing and projection matrix method, an improved projection matrix decoherence algorithm is proposed. Experimental results show that this algorithm is superior to the front-rear space smooth algorithm, and the amount of calculation is smaller.3. For the DOA estimation of independent source of coherent sources coexistence signal model, using differential technology and temporal smoothing algorithm, a difference time smoothing algorithm is proposed. This algorithm is superior to the classical decoherence algorithms in mixed signal model, and the array elements use ratio is higher.