Vehicle-Mounted Intelligent Surface for Cooperative Localization in Cellular Networks

Due to the continuous increase in the communication frequency band of the sixth-generation (6G) wireless communication network, it can potentially provide positioning services for applications such as intelligent transportation. However, limited radar cross sections of vehicles may result in too weak echo signals and incorrect measurement associations, especially for base stations (BSs) with low transmit power. To tackle this issue, we propose a novel cooperative localization scheme in cellular networks, where the controllable signal reflection capabilities of intelligent reflecting surfaces (IRSs) are utilized to facilitate target association. Specifically, IRSs are mounted on the vehicles to enhance the echo signals towards the associated BS while reducing interference with other BSs. We aim to minimize the maximum Cramer-Rao lower bound (CRLB) of location errors by jointly optimizing target association, IRS phase shifts, and dwell time. However, solving this CRLB optimization problem is non-trivial due to the non-convex objective function. To this end, we transform the problem for single-vehicle localization into a monotonic optimization problem, thereby optimally solving it through the Polyblock-based algorithm. For multi-vehicle localization, we propose a heuristic algorithm to efficiently optimize the target association and dwell time allocation. Finally, our simulation results verify that deploying IRSs on vehicles can effectively improve the resolution ability of multi-vehicle positioning and reduce the BS number requirements.