Adaptive NOMA/OMA in Short Packet Based Sensor Status Update Systems with Imperfect CSI

This paper studies the information freshness in short packet based sensor status update systems, where a control center constantly sends status packets to two Internet of Things (IoT) nodes to update the information, and the timeliness of status updates is characterized by age of information (AoI). To improve the AoI performance, the control center can adaptively select non-orthogonal multiple access (NOMA) or orthogonal multiple access (OMA) mode to send packets. Especially, considering that the IoT nodes cannot obtain accurate channel state information (CSI) in advance and the ideal successive interference cancellation (SIC) cannot be realized in actual short packet communications, we analyze the relationship of CSI estimation accuracy over time, and investigate the effect of imperfect CSI and SIC on status updates. To reduce the weighted AoI of the two nodes, we formulate a Markov Decision Process (MDP), where the control center can choose to send pilot and send packet via NOMA/OMA mode at each slot. We carefully analyze the state transitions in the MDP, transform the weighted AoI minimization problem into a linear programming problem, and then, the optimal scheme is obtained by solving the linear programming problem. The extensive simulation results show that the proposed adaptive NOMA/OMA scheme can significantly reduce the weighted AoI of the system.