Hybrid Noma-Based Resource Allocation for Multi-Access Edge Computing in Hetnets

This paper investigates the optimal offloading policy in heterogeneous networks where radio resources are provided, in both uplink and downlink, via two distinct types of BSs, namely Macro-cell Base Stations (MBSs) and Small-cell Base Stations (SBSs), in a multi-tier Multi-Access Edge Computing (MEC) -assisted scenario. Since the feasibility of the offloading problem is a function of radio connectivity in uplink and downlink, we propose to assign radio links using the flexible hybrid NOMA scheme that leverages both the limited interference of OMA as well as the faster data rates of NOMA. To this end, we formulate an optimization problem aiming to optimize the allocation of both radio and computation resources while minimizing the offloading energy across all users. The formulated problem is then tackled by means of decomposition and relation. The numerical results show that the Hybrid NOMA scheme balances subchannel allocation and power control to maintain high spectral efficiency and low offloading latency without the decoding overhead of Full NOMA (high interference) or the inefficiency of No NOMA (low subchannel reuse).