Research On Direction Of Arrival Tracking Algorithm Of Wireless Communication Signal

In order to further improve the capacity of system and the quality of data link in the mobile communication system, the high-accuracy DOA tracking based on space division multiplexing has become a research hotspot. The high-accuracy DOA tracking has received great attentions in several fields such as personal positioning service, mobile communications in battlefield, radar, sonar, earthquake and speech signal processing. Therefore, the high-accuracy DOA tracking is very important and has a broad prospect for development. Now, there is a large number of theory and technology research about high accuracy DOA tracking at home and abroad, but the traditional methods of DOA tracking still have many lilitations. So it will be significant to further improve the DOA tracking accuracy. This paper focuses on the research of high-accuracy DOA tracking algorithm. The major works are as follows:1. Firstly, a traditional method of DOA estimation based on MUSIC algorithm is improved to be applied to the DOA tracking. Then we compare it with another subspace-based method, URV-ESPRIT tracking method. The simulations show that the accuracy of MUSIC tracking method is much higher, but it cannot be used to track coherent sources and it requires a large amount of computation. While URV-ESPRIT, with small computational cost, also cannot process the coherent sources and the close trajectories.2. Then we examine two reduced complexity algorithms for subspace tracking. Projection Approximation Subspace Tracking(PASTd) and Fast Data Projection Method(FDPM) are presented and compared in terms of performance and efficiency. The analysis results indicate that PASTd has higher accuracy and less computational workload than MUSIC, but if the trajectoties are close to each other, its performance will deteriorate heavily. And the FDPM has high- resolution, however its performance will be reduced while the targets’ trajectoties become close.3. Based on the above studies, this paper using the sparse signal representation framework to track the DOA, then we propose a sparse-based DOA tracking technique. First, we propose a windowing method and a SVD-based method to discount the old data, mitigate the effects of noise and improve the performance. Then a general sparse Bayesian framework is introduced for DOA tracking. The sparse solutions are obtained to estimate the DOA’s change due to the target’s movement at each moment. We demonstrate the effectiveness of the proposed method with MUSIC and URV-ESPRIT. The simulation show that this method has better resolution, accuracy,imporoved robustness to noise, limitations in close trajectoties and correlation of the sources.This paper presents the comparision of the conventional subspace-based DOA tracking methods and fast subspace-based DOA tracking methods, and analyse their advantages and disadvantages. At the same time, this paper also proposes a sparse-based DOA tracking technique, and the simulation results show that those methods have more significance and broader developing prospect.