Adaptive Sensor Array Processing.
NAVAL UNDERSEA RESEARCH AND DEVELOPMENT CENTER SAN DIEGO CALIF
Pagination or Media Count:
The paper brings together some recent research results in array signal processing. Several theories for optimum processing of directional signals in noise fields with known statistics lead to processor structures which are amenable to adaptive parameter adjustment. Several steepest descent adaptive algorithms can be used for multichannel detection and estimation. The paper analyses the output signal-to-noise ratio of several power estimates using one of these algorithms. Assuming quasistationarity, the threshold directivity pattern is an efficient tool for evaluating the effects of slow changes in the signal and noise fields on the detection performance of any adaptive processor. Several examples illustrate the use of directivity patterns in the evaluation of array detection performance. Bounds on the excess mean-square error due to adaptation in a nonstationary environment show that the concept of quasi-stationarity does not imply strict stationarity. The complexity of the circuits implementing steepest descent adaptive algorithms can be reduced by averaging the gradient estimates, provided certain lower bounds on the number of bits in the quantized weights are used to avoid bias. This paper develops some analytical tools to assist the array processing engineer in deciding whether and how to design an adaptive array processor. Author