Array signal parameter estimation for CDMA technology (ASPECT)

Wireless communications have seen an unprecedented growth in the number of users and services offered over the last two decades. Wireless communications have expanded from the traditional analog cellular phone into the modern digital network capable of transmitting not only high quality voice, but also digital data. Along with the increase of capabilities has come a significant increase in complexity of signaling methods. The original systems that were deployed used conventional narrowband FM modulation; however, the third generation systems currently being deployed utilize direct sequence spread spectrum which is significantly more complex.; Smart antennas are an emerging technology for wireless systems that utilize a group of antennas in an array. The use of an array instead of a single antenna yields significant flexibility in the antenna system. One possible use for the antenna array is beamforming. Beamforming comprises the combining of various elements of the antenna array in such a way that the beam pattern can be changed. This allows the system to adjust the coverage area dynamically depending on the current traffic level. Another possible use for the antenna array is Direction Finding (DF) and geolocation. Geolocation is the automatic localization of a radio transmitter. This can be utilized to locate a user of a cell phone for emergency purposes (E911) or to aid in steering a beamformer. At the heart of these applications is Direction Finding (DF).; The increase in complexity of the signaling schemes has made the direction finding problem more difficult to solve. The third generation cellular systems being currently deployed are based on Code Division Multiple Access (CDMA) which utilizes direct sequence spread spectrum for the signaling scheme. One characteristic of these systems is a large number of co-channel users. Conventional direction finding techniques, such as MUSIC, are limited in the number of users' directions that can be estimated by an antenna array with a given number of antenna elements. In response to this limitation, a technique called the Pre/Post correlation technique has been developed which eases this limitation by utilizing some information of the signal structure. However, the Pre/Post correlation technique does not utilize all of the available signal information.; The technique developed in this dissertation is called Array Signal Parameter Estimation for CDMA Technology (ASPECT). The focus of this technique is to estimate the direction of arrival of CDMA users using all of the spread spectrum signal structure. Unlike the Pre/Post Correlation technique which uses only a single spreading code, this technique utilizes all of the spreading codes simultaneously. This allows the ASPECT technique to estimate the direction of arrivals in a robust fashion.; The ASPECT technique builds a second order covariance matrix for each despread signal. Since the base station knows the spreading codes used and the number of users a priori, a covariance matrix can be built for each user. A linear combination of these covariance matrixes is shown to estimate the array response vectors. The weighting values are determined using a linear least squares optimization method. In addition, a method that uses fourth order cumulant matrices instead of second order covariance matrices is presented. The effectiveness of the ASPECT technique is compared to the Pre/Post correlation and MUSIC techniques via a Monte Carlo simulation.