Target tracking over distributed sensor networks by polar measurements with time-varying bias

Abstract

This note concerns about the problem of target tracking over distributed sensors which measure range and azimuth with time-varying bias. First, the pseudo-measurements are generated to transform the biased nonlinear measurements into Cartesian coordinates. Then a modified two-stage filter is proposed to decouple the estimation of target state and sensor bias. Moreover, a distributed fusion algorithm for sensor network is derived based on local estimates. The effectiveness of the proposed filters are demonstrated by the Monte Carlo simulation results.