Mobile location in wideband CDMA communication systems

The objective of this research is to develop a mobile user location scheme for the future wideband code-division multiple access (CDMA) cellular networks. In a wireless environment, multipath propagation and line-of-sight (LOS) blockage present challenges for high accuracy mobile location.; In this thesis we present a hybrid location scheme for a wideband CDMA cellular system. In the system, the home base station (BS) measures the angle of arrival (AOA) of the transmitted signal from the mobile station (MS). Cross-correlators at the MS receiver are used to measure the time difference of arrival (TDOA) for different BSs. The forward-link TDOA measurements are relayed to the home BS, where the reverse-link AOA measurement for the MS signal is combined together to give a location estimate. The proposed scheme has the advantages of both the TDOA and AOA methods, and can avoid the drawbacks of the separate TDOA or AOA methods.; By introducing a linear form of the AOA equation, we develop two methods to solve the TDOA/AOA equations: the Taylor-series method based on the linear form of the AOA equation and the extended two-step least square (LS) approach. Simulation results demonstrate that the proposed hybrid TDOA/AOA location scheme gives much better location accuracy when the number of BSs is limited, and/or when the TDOA measurements have a relatively poor accuracy due to channel fading, non-line-of-sight (NLOS) propagation, etc.; Finally, the measurement bias produced in wireless location systems in the presence of NLOS error is analyzed for TDOA location. Based on the knowledge of NLOS error statistics, different NLOS mitigation algorithms are proposed. Using the TDOA LS residual, single NLOS BS can be identified with great accuracy. In the case of multiple NLOS BSs, an improved residual algorithm can effectively identify a small number of NLOS BSs. These residual algorithms require only the knowledge of Gaussian noise statistics. When more prior information is available, a maximum likelihood (ML) location estimator utilizing empirical databases can be used for improved accuracy. Simulation results demonstrate that an accurate location estimate is possible even in severe NLOS conditions.