Resource Orchestration in Interference-Limited Small Cell Networks: A Contract-Theoretic Approach

Abstract

Recently, non-orthogonal multiple access (NOMA) small cell networks (SCNs) have been studied to meet the stringent requirements for spectral efficiency and massive connectivity in the emerging 5G networks. This paper aims at addressing the overall resource orchestration issue in 5G SCNs, by considering the problem of joint user to small cell association and uplink power allocation, employing NOMA technology. In particular, the power allocation is performed under an incomplete information scenario, where the users’ channel conditions are unknown to the small cell. To treat this issue in an effective manner, contract theory is adopted in order to incentivize each user to select the power level that optimizes its own utility, while each small-cell base station (SBS) rewards them inversely proportionally to their respective sensed interference. The proposed framework is complemented by a distributed user-cell association mechanism based on reinforcement learning (RL). Indicative numerical results are provided to validate the operation and effectiveness of the proposed contract-theoretic approach.