Optimal Control of Status Updates in a Multiple Access Channel with Stability Constraints
share
In this work, we study peak-age-optimal scheduling with stability constraints in a multiple access channel (MAC) with two heterogeneous source nodes transmitting to a common destination. The first node has randomly arriving data packets and it is connected to a power grid. Another energy harvesting (EH) sensor node monitors the environment and sends status updates to the destination. We consider a stationary scheduling policy with random transmission decisions based on some probability distributions, and a dynamic policy where the decisions are made based on the observed network states in each time slot. With the stationary policy, we derive the average peak age of information (AoI) of the EH node and the stability condition of the grid-connected node, and optimize the transmission probabilities by minimizing the average peak age with stability constraint. The dynamic policy is developed by using Lyapunov optimization theory with the help of virtual queues. Simulation results show that the dynamic scheduling policy achieves lower average peak AoI compared to the stationary one, especially when the destination node have low multi-packet reception (MPR) capability. However, this advantage comes at the cost of frequent exchange of information between the nodes due to centralized scheduling.
READ FULL TEXT