New blind beamforming and moving target tracking techniques guided by the matrix-pencil/CDMA DOA algorithm
Said E. El‐Khamy, Ali M. Gaballa, H.E. Rachwan · 2005
The least squares constant modulus algorithm, LS-CMA, array blind beamformer has the main drawback of typically capturing the signal that has the greatest power among the received cochannel users. Hence, the capture process should be guided by controlling the initial beam steering. On the other hand, the matrix pencil (MP) algorithm is used to estimate the direction of arrival of all users incident in the array antenna. It uses a single snapshot to estimate the DOA and is not affected by correlation sources problem. Also, the matrix pencil algorithm when combined with CDMA is able to separate the DOA of the different users based on the orthogonality of the used spreading codes. The first presented technique in this paper is to use the MP-CDMA to blindly estimate the initial beam steering of the LS-CMA array, this combination allows to steer the initial beam of LS-CMA array antenna toward the desired user's direction. Therefore the desired user will have a dominant power at the array output and LSCMA will capture it. This technique is called MP-CDMA guided LS-CMA algorithm. The second presented technique to use the MP-CDMA combined with the minimum square (MS) beamforming for moving target (user) detection and tracking. The MP-CDMA is used to blindly estimate the steering matrix response of all the incident signals on the array at every snapshot and estimates the update array weight using MS beamforming. Computer simulations show the high efficiency of the proposed MPCDMA-LSCMA in separating CDMA signals of co-channel users. Also, we show that the MP-CDMA combined with MS beamforming can be efficiently used for tracking a moving user operating with other stationary users