Spectral-Energy EfficiencyTradeoff Enhancement: an Optimal Resource Allocation Framework for 5G Underlay Cognitive Radio Network

Abstract

This work considers the crucial problem of Spectral Efficiency (SE) - Energy Efficiency (EE) tradeoff enhancement in a 5G underlay Cognitive Radio Network (CRN). The SE-EE tradeoff problem is initially formulated as a Multi-Objective Optimization (MOO) problem for simultaneously maximizing the contradicting objectives of SE and EE. The tradeoff enhancement, being a Mixed Integer Non-Linear Programming Problem (MINLP), is difficult to solve. We propose a novel epsilon-constraint based linear programming resource allocation framework, which converts the SE-EE tradeoff problem to a single objective mixed integer linear optimization problem, solvable using standard linear programming techniques, providing optimal solutions. Simulation results are presented in terms of Resource Efficiency (RE), a metric for evaluating the SE-EE tradeoff. Simulation results show that high RE can be achieved at lower power budgets and it reduces as the power budget increases. Optimal RE at higher power budgets reduces for increasing epsilon, and interference temperature. Also, optimal RE increases for increasing number of subcarriers for the same 5G numerology. System design is performed adhering to 5G release 15 TR 138901.