Research For DOA Estimation Based On Propagator Method

Spatial spectrum estimation was an important research direction of array signal processing. High-resolution direction estimation of multiple sources in different directions in space could be achieved simultaneously. The existing algorithm of spatial spectrum estimation generally based on complex computation. Therefore, to reduce the computational complexity, enhance the performance in real-time and improve algorithm performance under the conditions of low SNR would be important parts to promote the wider use of spatial spectrum estimation techniques. In addition, a number of practical applications, the number of signal sources is greater than degrees of freedom of array, which is another key factor constrained practical application of spatial spectrum estimation algorithm.The main work of this paper was divided into two parts. First of all, A new method based on the fourth order cumulant which perform well on array extension was proposed to improve the utilization of array elements. The massive redundant data of the fourth-order cumulant matrix in uniform linear arrays can be eliminated by the application of minimum redundancy linear array.So a better capacity of array expansion can be obtained. Meanwhile, the application of propagator algorithm was extended to the fourth-order cumulant. The complex calculations which brings by eigen-decomposition was avoided, while the low-SNR flexibility of propagator method was also improved by the natural blind to Gaussian noise of the fourth-order cumulant. Simulation results show that the new algorithm has a strong capability, of array expansion, and has obvious advantages in low SNR conditions especially. The performance in array extension ability and computational complexity of the method is illuminated by computer simulations. In addition, two improved methods for propagator method of two-dimensional DOA estimation under the conditions of low SNR were proposed in this paper. The first one is based on the rearrangement of data-conjugate. An improved propagator method for two-dimensional DOA estimation based on two parallel uniform linear array is derived. The received data was made into conjugate data and rearranged, which is equivalent to a forward and backward smoothing. The performance of algorithm can be improved greatly in a limited number of snapshots and low SNR conditions. The second one is based on three parallel minimum-redundancy linear arrays, which use less number of array elements to obtain a larger array of effective aperture. The estimation performance in low SNR can also be improved. Computer simulation results show that the advantage of low-complexity was kept in both two algorithms above, while strong adaptability of low SNR and robustness were obtained concurrently.