Network State Sensing Assisted Resource Allocation for Grant Free Multiple Access Systems with Users of Heterogeneous Delay Tolerance

In the design of grant-free multiple access (GFMA) mechanisms, the number of active users (UEs) with heterogeneous delay tolerance is commonly assumed to be known, which is less practical in real-world implementations. In this paper, we propose a network state sensing assisted resource allocation algorithm for the GFMA systems, in which both the delay-sensitive unmanned aerial vehicles (UAVs) and the delay-tolerant terrestrial UEs simultaneously access to the base stations (BSs). Taking the limited radio resources and the practical assumption that the number of active UEs is not available at the BSs into account, we first design a Bayesian-based estimation algorithm that can determine the number of delay-sensitive UAVs and delay-tolerant terrestrial UEs in real-time. Based on the estimated result, we further develop a dynamic algorithm to judiciously allocate the limited radio resources according to the weights and number between the UAVs and the terrestrial UEs. Moreover, we dynamically adjust the access class barring factors to ensure the access delay requirements of the UEs. Through numerical results, we show how our proposed algorithm can achieve almost the same performance as that of the ideal algorithm assuming perfect information on the number of active UEs is available.