Resource Allocation in Virtualized CoMP-NOMA HetNets: Multi-Connectivity for Joint Transmission
share
In this work, we design a generalized joint transmission coordinated multi-point (JT-CoMP)-non-orthogonal multiple access (NOMA) model for a virtualized multi-infrastructure network. In this model, all users can benefit from multiple joint transmissions of CoMP thanks to the multi-connectivity opportunity provided by wireless network virtualization (WNV). We propose an unlimited NOMA clustering (UNC) scheme, where the order of NOMA clusters is the maximum possible value (called global NOMA cluster). We show that UNC provides the maximum overall spectral efficiency in CoMP-NOMA with maximum successful interference cancellation (SIC) complexity at users. To strike a balance between spectral efficiency and SIC complexity, we propose a limited NOMA clustering (LNC), where the SIC is performed to a subset of the global NOMA cluster sets. We formulate the problem of joint power allocation and user association in UNC and LNC such that CoMP scheduling and NOMA clustering are determined by the user association policy. Then, one globally and one locally optimal solutions are proposed for each problem based on mixed-integer monotonic optimization and sequential programming, respectively. Numerical assessments reveal that WNV and LNC improves users sum-rate and reduces users SIC complexity up to 65% and 45% compared to non-virtualized CoMP-NOMA and UNC, respectively.
READ FULL TEXT