摘要
The proliferation of time-critical IoT applications necessitates robust status update networks for real-time monitoring and autonomous control. To ensure stringent information freshness under total budget constraints, this work investigates the offline site planning problem for reconfigurable intelligent surface (RIS)-assisted networks. By approximating the average peak Age of Information (AoI) using an effective service rate, we transform the original min-max AoI objective into a tractable max-min rate optimization framework. The resulting mixed-integer linear programming (MILP) formulation jointly optimizes RIS locations, sizes, and long-term time allocation. To overcome the NP-hard combinatorial explosion, we propose G-MSR, a novel two-stage heuristic that strategically decouples the discrete deployment search from the continuous time allocation. Extensive simulations demonstrate that G-MSR achieves near-optimal solutions across diverse network scenarios while strategically reducing the computational overhead to a tractable polynomial bound, establishing it as a scalable tool for large-scale deployments.