Rapid Distance Estimation to Uncooperative Radars in Electronic Warfare Using Optimization and Transformer-Based Segmentation

While being tracked by an uncooperative radar, calculating the distance to the radar becomes a vital subject for electronic warfare (EW) systems. In most scenarios, the calculation is required to be immediate so that the EW system can respond quickly and dodge a possible threat. In this study, we show that a single moving EW system can find its distance to an uncooperative tracking radar in the order of seconds. We develop two different approaches utilizing non-convex optimization and transformer-based image segmentation methods. The optimization methods work directly on the time of arrival measurements of the received pulses. Moreover, we propose a pre-processing method to produce two-dimensional input images of the segmentation models, in which the instantaneous distance to the radar appears as an empty corridor. The models aim to extract this corridor, ultimately revealing the distance information. Quantitative results show that both approaches can accurately estimate the distance to the radars.