Jointly Optimizing Age of Information and Energy Consumption in Double-IRS-Assisted Wireless Networks

Age of Information (AoI) is a crucial metric for data freshness in 5G ultrareliable low-latency communication systems. It is affected by various factors, such as the channel quality, the energy of the nodes, and so on. To overcome the severe path losses and environmental obstacles, intelligent reconfigurable surface (IRS) is adopted to reconfigure signal propagation environments, thereby mitigating the significant losses and creating additional signal reflection links. Different from the existing research on AoI in IRS-assisted wireless networks, which overlooks the impact of energy consumption on AoI and assumes ideal phase shift models, we investigate the entire process of data generation to arrival and propose a double-IRS cooperative scheme for data transmission with practical phase shift models in blocked wireless networks. The joint optimization problem of the long-term average AoI and node energy consumption is formulated. Given the nonconvexity of the joint optimization problem, we decompose it into an outer data transmission scheduling problem and an inner double-IRS phase shift optimization using a two-layer optimization framework. Then, we use the deep Q-network algorithm to solve the outer problem. For the inner problem, we propose a penalty-based alternating optimization (PB-AO) algorithm to solve the coupling between double-IRS phase shifts and the nonconvex constraints of practical phase shift models. The extensive simulation results verify the effectiveness of the proposed PB-AO algorithm in optimizing double-IRS phase shifts under the practical phase shift model and demonstrate the superiority of our double-IRS cooperative data transmission scheme compared with other benchmark ones.