Maximization of Users’ Fairness in an Imbalanced-NOMA Network scenario with More Far-Users, by means of Multiple Near-Field Relays

Abstract

The development of 5G networks is currently prominent in the mobile network industry. This is due to the need for increased network performances, such as more capacity and reliability (users’ fairness); which the current 4G networks fall short of delivering. Non-Orthogonal Multiple Access (NOMA) technology has been identified by others, as key elements for the realization of 5G networks; thus, the design of NOMA networks has recently gained popularity. NOMA technology implies serving 2 or more users in the same frequency band and discriminating their respective information by assigning different power levels to each (PDNOMA). In the case of 2 users per frequency band, a base-station with N-antennas must serve 2N users at the time. Initial works in the NOMA-networks’ design often assumed that the 2Nusers to be served consist of N-far field users and N-near field users. This scenario is referred to as the “balanced-NOMA” scenario. Initial power-allocation algorithms (Initial FM-PAA) have been proposed to maximize users’ fairness of NOMA networks in this scenario, and have yielded relatively good results. However, there can be a case where there are more far-users than near-users in the 2N set to be served; it is referred to as an “imbalanced NOMA” scenario. In this case, if the “initial FM-PAA” is used straight, it will only serve possible pairs, leaving many far-users unserved. This will result in very poor user fairness for the network. To address this problem, first, an “intermediate PAA”, which executes NOMA combined with OMA, was designed, implemented and tested. The algorithm consists of an inter-beam powersharing stage, to distribute the base station’s power across respective antennas; and an intrabeam power-sharing stage, only applicable to the NOMA pairs formulated. Both stages employed the “OCTR-ratios convergence concept”. The results indicated that the proposed “intermediate-PAA” considerably improves the fairness of the imbalanced-NOMA network scenario; compared to when the “initial FM-PAA” is used straight. However, since it does not serve all users, it therefore constitutes an intermediate solution to the problem stated. Furthermore, the research proposed an “advanced-PAA” solution, to completely address the problem. The solution consists of placing relays in the near-field of the base station. Each relay serves as the near-user to one of the unpaired far-users, and it will be served with the information intended for the other unpaired far-user. This turns the system into a perfectly balanced NOMA scenario. Then, a power-allocation algorithm, which combines an “initial FMPAA”, and a “relay management system”, was designed, implemented and tested. The “initial FM-PAA” was designed based on the “OCTR-ratios convergence” concept. The “relaymanagement system” was based on the “decode and forward” concept. The results demonstrated that the proposed “advanced PAA” maximizes the fairness of the imbalanced- NOMA network scenario; and as such, outshines, both the “intermediate PAA” and the “initial FM-PAA” used straight. Therefore, it provides an optimal solution to the stated problem.