GPR-Based Based Blockage Prediction and Optimal Path Selection in IRS and Relay-Assisted V2X Communication

The paper presents a framework of an IRS (intelligent reflecting surface) and Relay-assisted V2X(Vehicle to everything) communication, as vehicular communication systems face various challenges and issues due to dynamic environment and signal blockages. In this work, advanced predictive models like Gaussian Process Regression (GPR) and Gated Recurrent Unit (GRU)-based estimation are utilized to evaluate IRS and Relay-assisted system. The intention is to enhance the coverage and efficiency of V2X with the help of some of the recent technologies identified such as Intelligent Reflecting Surface (IRS) and Relay-Assisted Communication (RAC). This paper presents a comparative analysis of IRS-assisted and DF relay-assisted V2X communication through experimental results. GPR and GRU models serve as a basis for blockage prediction, signal-to-noise ratio (SNR) estimation, and performance analysis by utilizing data rates and best path selection. Results reveal that IRS-assisted communication can strengthen the channel link and ensure effective blockage alleviation even though it exhibits marginally larger prediction errors compared to DF relays. The achieved results show potential of IRS-assisted systems which is more reliable and capable to enhance V2X communication.