Age of Incorrect Information Minimization for Timely Status Updates in Satellite-based IoT

In satellite-based Internet of Things (S-IoT) applications such as climate change and natural disaster monitoring, timely status updates are very important. In this paper, we utilize the content-aware performance metric, named age of incorrect information (AoII) to design an AoII minimization power allocation scheme for multiple user equipments (UEs) transmit their short-packet status updates via uplink rate-splitting multiple access (RSMA) over shadowed-Rician (SR) fading channels. Specifically, we formulate a long-term average AoII minimization optimization problem with average and peak power constraints. By means of Lyapunov optimization framework, the original problem is transformed into a problem that minimizes the upper bound of Lyapunov drift plus the penalty term, and then deep reinforcement learning (DRL) algorithm named multi-agent twin delayed deep deterministic policy gradient (MATD3) is used to solve our AoII minimization power allocation (AMPA) scheme. Simulation results show that our proposed AMPA scheme can achieve the optimal long-term average AoII performance than state-of-the-art schemes.