Research On DOA Estimation And Application Based On TD-SCDMA Network

TD-SCDMA network adopts the new techniques of uplink synchronization and Uniform Circular Array (UCA) smart antenna, so the serving base station can obtain the Time of Arrival (TOA) and Direction-Of-Arrival (DOA) of mobile users' signals conveniently. Based on those parameters we can obtain the mobile users' position by only one base station, and the precise DOA estimation is the crucial technique. However, those classical DOA estimation algorithms based on Uniform Linear Array (ULA), such as Multiple Signal Classification algorithm (MUSIC) and Estimation of Signal Parameters via Rotational Invariance Technique algorithm (ESPRIT) can't be used in UCA directly, how to apply these classical algorithms to UCA is a important research item.Based on the analysis to research status of DOA estimation techniques, wireless location techniques and the characteristic of the TD-SCDMA network, subspace-based DOA estimation and its application are determined to be the main research item in this thesis. In order to estimate the users' signal DOA precisely, firstly this thesis pays more attention to the research of two kinds of subspace-based algorithms known as MUSIC and ESPRIT, besides, several improvement suggestions to these algorithms are proposed. As MUSIC algorithm can't distinguish the DOA of highly correlated multipath signals, spatial smoothing is used to de-correlate the signals in the ULA, but the spatial smoothing techniques don't fit the UCA. In order to apply spatial smoothing techniques to UCA, We convert the array manifold from UCA to virtual ULA by Pretreatment techniques, and propose the mode-space SS MUSIC algorithm which can de-correlate the correlated signals in UCA. For solving the fluctuation of signal subspace caused by noise, two kinds of TLS-ESPRIT are proposed. In order to apply those ESPRIT algorithms to UCA, the mode-space algorithms of UCA are introduced. Based on proper mathematic model, the performance of those algorithms is simulated and compared in detail. Finally, based on above research, the project of applying the DOA estimation to TD-SCDMA location systems is proposed. Based on the foundation of proper channel model, the location performance of the project is simulated by matlab in detail, and the simulation results verify the applicability of the project.