AoI Minimization for Grant-Free Massive Access with Short Packets using Mean-Field Games

Abstract

Grant-free (GF) access, where channels are accessed without undergoing assignment through a handshake process, is a promising solution to support the massive connectivity for IoT networks. In this paper, we consider uplink GF massive access for an IoT network. IoT devices generate short packets and transmit the generated packets by GF non-orthogonal multiple access (NOMA) communications. To keep the information fresh, we first derive the age of information (AoI) in the GF short-packet communications and then formulate the AoI minimization problem. However, the AoI minimization problem is challenging to solve since the number of users involved is large. To tackle this problem efficiently, we propose a mean-field evolutionary game-based scheme where the average behavior of the IoT nodes will be considered rather than their individual behavior to reduce the complexity. Simulation results verify the effectiveness of the proposed mean-field evolutionary game-based algorithm.