Application of Robust statistics to target bearing estimationfor split arrays in active SONAR
Yang Yun-chuan · Shengxue jishu · 2004
The normal mode (least square producers) loses efficiency drastically when the length of digital signal processing does not match the length of echo, or the SNR is not a constant in target bearing estimation. Indeed, a single outlier can have an arbitrarily large effect on the estimate. It causes significant errors in regression coefficients. This paper analyzes the reasons and gives a time-bearing model. Robust statistics and Huber-estimator are introduced. A modified Huber estimator presented and its application introduced. Simulation results show that the method can provide better and robust effect on bearing estimation..