Study On DOA Estimation Algorithms Of Array Signal

Array signal processing technique comes from signal processing. In the fields of radar, sonar, radio sky telescope, GPS, wireless communications, medical and so on, it is a critical technology to service people’s work and life.Array signal processing includes two main directions:Beam forming (BF), direction of arrival (DOA) Estimation. Wherein the content of this paper is DOA. Spatial spectrum estimation, namely the estimation of DOA, uses the principle of an array antenna receiving signals in a noisy environment to estimate the incident direction of relative array antenna.This paper discusses the relevant knowledge array signal processing, including the distinction between narrow band and wide band signal, the source of coherent and incoherent, and model noise and the like. And analyzes four antenna model, including the uniform linear array (ULA), uniform circular array (UCA), prime array, and nested arrays, among which the prime nesting arrays and array are sparse matrix. About DOA research algorithms, first, the paper cited three traditional DOA estimation algorithm, namely conventional beam forming (CBF) algorithm, Capon algorithm, the traditional MUSIC algorithm. At the same time, it demonstrates its specific principles and gives the DOA Figure estimate simulation of each algorithm under the same condition, including spatial spectral error analysis and other indicators, and comprehensive compare of algorithm performances. Then, direct at a kind of particular signal, coherent signal, analyze two types of DOA decoherent estimation algorithm, including spatial smoothing algorithm, and improved MUSIC algorithm.In order to reduce the overall cost of the antenna, antenna arrays need to be thinned. This technology is a way to remove some of the array element from the existing uniform array, or to recombine to sparse array of non-equidistant according to certain rules, to achieve an array rarefaction. Not only sparse array meet the requirement of angle measurement accuracy, but also the whole system cost minimizes.This paper focuses on the fifth chapter, which analysis DOA estimation algorithm of two forms of sparse array Co-prime arrays and Nested arrays, including DOA estimation methods based on the Co-prime arrays:spatial smoothing algorithm, sparse reconstruction convex optimization algorithm, Lasso convex optimization algorithm, and DOA estimation method based on the Nested arrays:spatial smoothing algorithm. At the same time, by simulation, the following conclusions can be drawn. As the three DOA estimation methods base on the Co-prime arrays, spatial smoothing algorithm need the number of targets as priori knowledge, sparse reconstruction convex optimization algorithm spend a lot of time, and there are many false peaks in the spatial spectrum of Lasso convex optimization algorithm. In view of the above problems, a modified algorithm base on Lasso convex optimization algorithm is proposed. Simulation results indicate that the algorithm using the HTSWS and power estimation enhancement could overcome above problems of the three algorithm.