Research On Wideband Source DOA Estimation Algorithms

Spatial spectrum estimation is an important branch of the array signal processing. The main purpose of spatial spectrum estimation is estimating the parameters of signal, such as carrier frequency, direction of arrival (DOA) and so on. DOA (DOA) estimation algorithm has found numerous applications in radar, sonar, and microphone array speech processing, more recently, in wireless communications.So far, narrowband signals DOA (DOA) estimation algorithm has been more deeply researched. However, wideband signal has encounted in many important applications such as acoustic location, underwater signal processing, etc. This paper discusses some wideband subspace DOA estimation algorithm for uniform linear array (ULA). The main work is summarized as follows:1. Several narrow-band DOA estimation algorithm such as Beam-forming, linear prediction, MUSIC, ESPRIT are discussed. MSE of those algorithms are compare with CRB.2. CRB of wideband DOA estimation is briefed derived and researched by simulation. Studies have shown that single source CRB is inversely proportional to the SNR as well as azimuth, and two sources CRB is also inversely proportional to SNR, but increasing with between angles of sources.Wideband DOA estimation algorithm such as ISM,CSM are exactly studied. Improved ISM using matrix reconstruction method is proposed. Simulations results show that under certain conditions, the improved ISM algorithm can effect distinguish coherent wideband signals. Moreover, several wideband focusing algorithm are researched using simulation. The results show that wideband focusing algorithm has better resolution ability and estimated accuracy than that of ISM algorithm.