The Studying Of Signal Direction Of Arrival (DOA) Tracking Algorithm In Smart Antenna

Since3G standard TD-SCDMA is proposed in china, the smart antennatechnology has absorbed the interest of the scholars all over the world. Thistechnology has very important significance in improving spectrum utilization,enlarging channel capacity and increasing the distance covered by communicationsystem. The research on smart antenna involves many techniques and algorithm,where the Direction of Arrival (DOA) estimates algorithm plays an importance role.The traditional high-resolution DOA estimation algorithm is proposed byeigen-decompositing the covariance matrix of array output signal, therefore, thealgorithm is a batch method, which contains a mount of computation. Thus thisalgorithm can not be used to track the DOA of time-varying signal. Moreover, theexisting tracking algorithm has contradiction in DOA tracking error and trackingspeed, thus the topic of this paper is to come up with new algorithms, which performa low DOA tracking error, and achieve faster DOA tracking.The components of the smart antenna system and the theories of DOA estimationfrom sampling signal are summarized in the paper. Meanwhile, the merits anddemerits analysis of various estimation algorithms are also introduced in detail,where the MUSIC and ESPRIT algorithm has the merits of super-resolutioncapability. However both of the method could not lack tracking the DOA of thesignal. In contrast, the PAST and FAPI algorithm can accomplish tracking the DOA,but could not resolve the conflict between the DOA tracking error and the trackingspeed.Based on the conclusion mentioned, a novel subspace updating algorithm withvariable forgetting factor is proposed in this paper, which can mitigate the conflictbetween the DOA tracking error and the tracking speed to a certain extent. On theother hand, combined with the phase mode motivate algorithms (such asUCA-ESPRIT, etc.), a new method in two-dimensional DOA tracking algorithm isalso presented, which achieves tracking azimuth and elevation angle synchronized, and exhibits more practical purposes. Finally, the simulation results verify thefeasibility and effectiveness on the two-dimensional DOA estimation.