Data-Driven Random Access Optimization in Multi-Cell IoT Networks Using NOMA

Non-orthogonal multiple access (NOMA) is a key technology to enable massive machine type communications (mMTC) in 5G networks and beyond. In this paper, NOMA applied improve the random efficiency high-density spatially-distributed multi-cell wireless IoT networks, where devices contend for accessing shared channel using an adaptive p-persistent slotted Aloha protocol. To capacity-optimal network, novel formulation of management proposed, which transmission probability each device tuned maximize geometric mean users' expected capacity. It shown that network optimization objective high dimensional mathematically intractable, yet it admits favourable mathematical properties design efficient data-driven algorithmic solutions do not require priori knowledge model or topology. A centralized model-based algorithm scalable distributed model-free algorithm, are proposed optimally tune probabilities attain maximum The convergence algorithms optimal solution further established based on convex game-theoretic analysis. Extensive simulations demonstrate merits efficacy algorithms.