Propagation of Heterogeneous Receiver Self-Positioning Uncertainty to Target Localisation and Tracking in Airborne Multistatic Passive Radar

Abstract

This work investigates an airborne (UAV-based) multistatic passive radar using signals of opportunity in the context of a GNSS-denied environment. Surveyed works on multistatic passive radars either assume zero receiver positioning error, or otherwise ignore its heterogeneity across receivers, which we argue is expected in GNSS-denied navigation due to varying conditions. Rui and Ho’s error model accommodates heterogeneous per-receiver errors but is evaluated only for the IID case. We adapt this model and study the heterogeneous regime it permits with a weighted least-squares (WLS) solver, and operationalise their deferred drifting-receiver case via an EKF tracker with per-receiver drift state. We run a Monte Carlo simulation with 10,000 trials per configuration varying the number of receivers and receiver positioning error parameters to compare a per-receiver-error-aware solver against a solver assuming IID error across receivers, and a solver assuming no self-positioning error at all. The findings show that, relative to the IID solver, the per-receiver-aware solver reduces the median target-position error by ∼13% in snapshot WLS (at N =10, σRx =5 m) and the median steady-state target-position RMSE by ∼28% in EKF tracking (at N =10, σdrift=0.1 m/√s). This suggests that in a UAV-based multistatic passive radar, per-receiver self-positioning error is a non-negligible parameter to consider in design.