QoE-oriented resource allocation for dense cloud NOMA smallcell networks

In this paper, we investigate the resource allocation for dense cloud non-orthogonal multiple access smallcell networks (NOMA SCN), aimed at maximizing the users' quality of service (QoE). First, we construct a directed hypergraph to model the complex inter-interference relationship for cloud NOMA SCN. Then, we formulate the QoE-oriented channel allocation and user pairing problem in NOMA SCN as a local cooperation game. The game is proved to be an exact potential game. Moreover, the optimal pure strategy Nash equilibrium (PNE) in the proposed game can maximize the network QoE level. To achieve the optimal PNE in the proposed game, we redesign a directed-hypergraph-based multi-agent learning algorithm, which allows multiple non-coupled agents in directed hypergraph to simultaneously update their actions. Finally, simulation results are presented to validate the proposed learning scheme.

Automated summary

This paper investigates resource allocation for dense cloud non-orthogonal multiple access smallcell networks (NOMA SCN) with the aim of maximizing users' quality of service (QoE). A directed hypergraph is constructed to model the complex inter-interference relationship in cloud NOMA SCN. The QoE-oriented channel allocation and user pairing problem is formulated as a local cooperation game, which is proven to be an exact potential game. An optimal pure strategy Nash equilibrium (PNE) is identified in the proposed game to maximize network QoE level. A directed-hypergraph-based multi-agent learning algorithm is redesigned to achieve the optimal PNE. Simulation results are presented to validate the proposed learning scheme.