Spectrum Allocation for UAV-Aided Relative Localization of Ground Vehicles

High-accuracy localization capability is important for applications of autonomous vehicles. For tasks in automatic drive such as overtaking, relative position errors are often more concerned than absolute position errors. In this paper, a spectrum allocation scheme is proposed for relative localization of the vehicles aided by the unmanned aerial vehicles. The squared position error bounds (SPEBs) for both absolute and relative localization are derived where subspace projection is used for the relative SPEB. Then, the spectrum allocation problem is formulated for relative localization. We show that the optimal solution of the spectrum allocation for relative localization can be obtained by semi-definite programs. Numerical results show that optimal spectrum allocation outperforms the uniform allocation by 50% in the root relative SPEB.