Enhanced RIS-assisted vehicular network with TDMA and Bayesian Compressive Sensing-based channel estimation

In recent years, vehicular communication networks have become increasingly critical for intelligent transportation systems and autonomous driving applications. However, traditional vehicular networks face significant challenges in achieving reliable high-throughput communication, particularly for vehicles at the network edge or in non-line-of-sight scenarios. While Reconfigurable Intelligent Surface (RIS) technology offers promising solutions through programmable signal reflections, the joint optimization of RIS configuration and resource allocation in dynamic vehicular environments remains a complex and open challenge. In this paper, we propose an RIS-assisted uplink multi-input single-output (MISO) vehicular network communication system, where vehicle sensors transmit the collected data to roadside units (RSUs) in their specific time slots. To enhance transmission efficiency and reliability, we employ an adaptive Time Division Multiple Access (TDMA) scheme, which assigns dedicated time slots to each vehicle, thereby avoiding signal collisions and improving spectrum utilization. Furthermore, to address the channel estimation challenge in mobile scenarios, we develop a practical and efficient channel estimation framework based on Bayesian Compressive Sensing (BCS). Specifically, to leverage the inherent sparsity in the channel structure, our approach minimizes pilot overhead while enabling accurate and efficient recovery of the channel state information (CSI) in both direct and RIS-assisted paths under Rician fading conditions. To maximize the system throughput through the joint optimization of RIS phase shifts, power allocation, and time slots, we utilize the Block Coordinate Descent (BCD) algorithm to solve this non-convex optimization problem. The numerical results demonstrate that the proposed BCS-based method significantly enhances channel estimation accuracy and system throughput compared to other state-of-the-art approaches.