ISAC-Enabled Near-Field High-Mobility UAV Localization: Analysis, Bounds and Algorithm

Reliable localization of high-mobility unmanned aerial vehicles (UAVs) is essential for various services in the lowaltitude economy. Future mobile networks offer an economical solution for wide-area UAV localization via integrated sensing and communication (ISAC) capabilities. Operating in the near-field region due to larger antenna arrays and higher frequencies, these networks are subject to spherical wave propagation, which induces intrinsic coupling among range, angle, and Doppler shifts. Moreover, this coupling is compounded by the high mobility of UAVs, a critical aspect that existing signal models for nearfield localization may not adequately address. In this paper, we analyze the performance degradation in target localization caused by such mobility-induced model misspecification, and propose a localization algorithm for near-field high-mobility UAV target. Specifically, the fundamental limits on achievable parameter estimation accuracy in the presence of model inaccuracies are quantified using the misspecified Cramer-Rao bound (MCRB). Building upon this analysis, we develop a two-stage near-field localization approach to effectively mitigate the parameter coupling exacerbated by high mobility. Initial parameter estimates are efficiently obtained through tensor decomposition, and subsequently refined via quasi-Newton optimization. Simulation results validate the superior localization and velocity estimation accuracy of the proposed algorithm, aligning well with the theoretical performance bounds