A variable length linear array algorithm for smart antenna systems

Smart Antenna technology has numerous advantages that improve wireless communications. For example, interference rejection, capacity and range increases, and the ability to track both stationary and mobile users. A key research area is the development of adaptive algorithms,which stems from improvements in processing power. For example, algorithms such as the Least Mean Squares can be deployed successfully to allow the antenna system to track mobile users. Furthermore, the use of adaptive algorithms for real world applications has also received greater notice. These algorithms, known as decision directed adaptive algorithms utilise information from the surrounding wireless environment to perform adaptation. Henceforth, the overall objective of this thesis is to present a new decision directed adaptive algorithm. A new variable length linear array algorithm is presented which uses the beam width and angle difference between a desired signal and interfering signal as a decision criteria. In doing so, the number of radiating elements is determined and activated, hence allowing an antenna array to vary its size based on the scenario. This thesis presents two forms of this algorithm. The first approach uses the LMS algorithm to assist in the calculationof weight coefficients once the number of elements has been determined. The second approach uses partial optimisation to carry out the responsibility of the adaptation process. In this instance, weight coefficients of large arrays are applied to sub-arrays through the use of a look up table. This approach is shown to maintain radiation pattern characteristics such as the beam width.