Research On The Direction Of Arrival Estimation Of CDMA Signals

Based on the theory of spread spectrum, the communication method of Code Division Multiple Access has the advantages of anti-interference and working in low power spectrum density, so it has become the key technology of the third generation mobile communication system. In order to improve the quality of systematic performance, mobile communication system is bound to adopt the array antenna technology, and the spatial spectrum estimation research in the technology is also the hot spot in the recent decade.Direction of arrival(DOA) estimation has been focus of research in array signal processing. Among different methods developed for DOA estimation, subspace methods have been found an outstanding prominence due to their high resolution and low complexity. However, in the complicated wireless communication system like CDMA, it is hard to use the subspace algorithm like MUSIC directly. As spreading code has correlation properties and good cross correlation properties, each transmitted user signal in CDMA system is distributed with a respective Pseudo random sequence. The paper would research on how to utilize the property to estimate DOA for CDMA signal. This paper is divided in three parts.Part one would first focus on DOA estimation in one-path situation. As MUSIC algorithm is a representative high resolution DOA estimation technology, the number of elements is required to be at least 1 more than that of incident signals. However, in CDMA system, as the number of users is too much, the limitation above is not appropriate. Therefore, in the situation of knowing the spreading code of users, code matched filter is adopted to preprocess the received signal, and MUSIC algorithm is used to angle estimating for each user. Then considering multi-path situation, multiple access interference can be inhibited to some extent after the decorrelating detector of multiuser detection is combined to code matched filter.Combined with DOA estimation algorithm and aiming at non-cooperative communication environment, the second part would estimate the DOA of CDMA signal in the situation that the receiver does not know the spreading code of the expected user. Firstly, pseudo code is estimated by using the method based on eigendecomposition. Then, after getting the spreading code, the DOA estimation of each user is got through the paralleled MUSIC algorithm which is based on code matched filter.Part three would introduce the joint angle and delay estimation. Because of the influence of multi-path, when the receiver is dealing with CDMA signal, time delay is also a very important influencing parameter. Two important conceptions, spatial array manifold and temporal array manifold in joint angle and delay are introduced. At last, based on TST-MUSIC joint angle time delay estimation algorithm for CDMA signal is described. By combining temporal filtering and spatial beam forming in the algorithm, the delay and angle estimation value in each relative path is got.