2-D VESPA Algorithm for Multiuser and Multipath DOA Estimation in DS-CDMA System

We propose an effective two dimensional (2-D) direction-of-arrival (DOA) estimation algorithm for the synchronous DS-CDMA systems in multi-user and multi-path channels. The algorithm can be applied in CDMA base station receiver with arbitrary array antenna. In order to effectively suppress multiple access interference (MAI), multiple parallel code-matched filters and a decorrelating multiuser detector are used to detect each path of desired signal for each mobile user. In the output of the detector, a cumulant-based 2-D Virtual ESPRIT Algorithm (VESPA) is employed to estimate each resolvable path direction-azimuth/elevation. In the method, 2-D DOA of each sub-path signal is accurately estimated with higher resolution than the previous methods because array aperture is extended by the use of higher order cumulants. It also has advantage of eliminating the spatially colored Gaussian noise and being available in arbitrary array geometry.