Array Calibration Technique In DOA Estimation

DOA (Direction of Arrival) estimation is an important research area of array signal processing with growing applications in engineering, such as radar, sonar, communication, biomedicine, etc. A number of highly accurate methods exist for solving this problem. Unfortunately they all require precise knowledge about the characteristics of the receiving antenna array. These characteristics, which include sensor locations, gain and phase response, and mutual coupling, are rarely perfectly known in real situations. So this dissertation aims at the development of array calibration and robust DOA estimation methods, based on different application background and real environment.A self-calibration algorithm is proposed, which can correct the sensor position errors (or gain and phase errors) in nonuniform noise fields. In this algorithm, the multi-dimensional self-calibration method is combined with the maximum likelihood (ML) estimation method of the colored noise covariance matrix. It can provide more accuate initial value to the subsequent iterative process. The simulation was conducted and it was indicated that in low SNR scenarios, this algorithm has good performance of calibration and DOA estimation.The problem of DOA estimation and array gain and phase calibration in the presence of multipath is studied. First, two decorrelation methods (spatial smoothing and mode space transform) are discussed, and their influences to the calibtation of array gain and phase are analyzed. Then based on the weighted subspace fitting (WSF) scheme, a self-calibration algorithm is proposed, which can correct the array gain and phase errors in the presence of multipath propagation. The first-order derivative and Hessian matrix of this method's cost function are derivated, and the alternating optimization and Gauss-Newton method are used to realize the multidimensional search process. This method's calibration and DOA estimation performance is illustrated through computer simulations.Regarding the situation that array gain/phase error and mutual coupling exit simultaneously, a joint calibration method is developed, which has superior calibrating range. By using the H~∞filter which is proposed in robust control theory to restrain the influences of array model errors and environment noise, the more accuate initial value can be provided to the subsequent joint iterative self-calibration process. This method has no need of any a priori knowledge about those errors and has higher resolution. So this method is easy to implement in practice.In the presence of array imperfection and mutual coupling, a wideband robust DOA estimation algorithm which based on the signals separation category is proposed. By using the matrix operators which have the memory and oblivion characteristics, this algorithm can separate the incident signals effectively and avoid the iterative process falling into the local extreme value. This method has no need of array calibration and measurement beforehand. So it has good potential in applications.