Bias Estimation of Multiple Radars by Quasi-Recursive Filtering

Abstract

This paper presents a method for estimating bias in radar measurements used in tracking aerospace targets. The bias-free and less noisy radar measurement is critical as it would provide correct information of target kinematics which is a necessity of the guidance law to generate the tracking commands. In a typical interception scenario, multiple radars are employed to detect and track target kinematics. Bias estimation is often difficult because of limited observability of sensor biases as there may not be a unique set of biases that explains the relative errors between measurements. The sensors involved may be dissimilar, and their corresponding bias parameters may differ in magnitude and type. The various sources of biases are sensor bias, alignment error (tilt error), and radar position uncertainty. Radar measurements are to be debiased and noise filtered effectively before they can be used singly or in conjunction with any onboard sensor for generating tracking commands. The formulation of bias estimator described here is designed for tracking systems, such as aircrafts, ships, with multiple sensors as radars, jam strobe detectors, GPS, ESM and wider variety of targets and sensors. The paper concisely states the algorithm which addresses the above mentioned problems and illustrates its performance capabilities through results obtained by applying it to a realistic ballistic target tracking scenario