Minimum Variance Estimators for Enemy Radar Localization by Unmanned Aerial Vehicles

Cooperative team of Unmanned Aerial Vehicles (UAV) is utilized for deceiving the radar network of enemy by generating a phantom track. So the enemy radar position needs to be estimated in advance by the UAV using Time Difference of Arrival (TDOA) estimation method. The radar pulses received by UAV are interrogated and delayed signals are sent back which essentially deceive the radar and gives an idea of a phantom trajectory to radar network. However the estimation of TDOA needs to be accurate and fast. Several approaches have been reported in the literature for estimation of TDOA signal. The system dynamics is nonlinear. Linearised time varying Kalman filter and Extended Kalman Filter (EKF) are some of the methods which are used for the purpose. However under uncertain and noisy environment, the estimation result is prone to be inaccurate. The variance of estimation error is high. The alternative design is required to take account of the noisy process. The radar position has been estimated by TDOA geolocation technique on utilizing Unscented Kalman Filter (UKF) and Particle Filter (PF) and the comparative evaluation of unconstrained filters (EKF, UKF, and PF) has been demonstrated on a typical radar network. This paper outlines the applied estimation schemes for on board estimation in UAV, showing overall improvement in the variance of its position estimates for the middle radar.