Two-Timescale Vehicle Association and Resource Management for C-V2X Networks

Short-timescale resource management for dense cellular-based vehicle-to-everything (C-V2X) networks inevitably incur heavy signaling overhead and system latency caused by frequent resource re-allocation. To reduce it, the most effective way is to decrease the frequency of resource allocation. But this will cause great mismatch between the quality-of-service (QoS) requirements and available resources, and further will degrade the efficiency of resources, and finally decrease the QoS of vehicles. To solve the problem and achieve a win-win situation, i.e., improving the system performance and decreasing the signaling overhead, this letter devotes itself to study a two-timescale optimization scheme by considering the mobility nature of users in C-V2X networks. The proposed scheme is in a hierarchical optimization architecture, which includes 1) long-timescale vehicle association and resource management, and 2) short-timescale power control. For the long-timescale, a matching-auction-based vehicle association, spectrum assignment and power control algorithm (MA-VASAPCA) is proposed. Given the long-timescale decision, we derive the closed-form expression for the power variable that satisfies the instantaneous QoS requirements of various services. Simulation results show that the proposed scheme can significantly enhance the system sum-rate, guarantee the QoS requirements of users and reduce the system signaling overhead.