Age of Information Driven Power Allocation with Sum-Power Constraint in Downlink Networks

Age of information (AoI), a metric for data freshness in status update systems, has garnered increasing attention in various applications. This paper focuses on dynamic power control to minimize the average AoI for all users in a power constrained downlink network, where each user has three selectable power levels in every time slot. By relaxing the instantaneous power constraint to a time-averaged power constraint, we decompose the Lagrangian minimization problem into multiple independent power allocation subproblems, where each corresponds to a specific user. We demonstrate that each subproblem possesses the full indexability property. By comparing the Whittle indices with the Lagrange multiplier, full indexability allows us to determine the optimal Lagrange multiplier, which effectively utilizes the available power budget. Based on this, we propose a low-complexity power allocation approach to solve the problem. Simulation results validate our analysis and show that the proposed approach outperforms benchmark policies.