DYNAMIC RESOURCE ALLOCATION OF EMBB/URLLC TRAFFIC IN 5G NR

Abstract

5G technology is intended to support three promising services with heterogeneous requirements: Ultra-Reliable and Low Latency Communication (uRLLC), enhanced Mobile Broadband (eMBB) and massive Machine Type Communication (mMTC). The presence of these services on the same network creates a challenging task of resource allocation to meet their diverse requirements. Given the critical nature of uRLLC applications, uRLLC traffic will always have the highest priority which causes a negative impact on the performance of other types of services. In this thesis, the problem of uRLLC/eMBB resource allocation is addressed. Sub-optimal and optimal solutions are proposed. Heuristic scheduling algorithms are utilized in the sub-optimal approach, providing a low complexity solution to the problem. A knapsack inspired punctured resource allocation algorithm is proposed in which the channel quality of both eMBB and uRLLC UEs are considered at each time slot to make the best Resource Block (RB) selection for puncturing in a way that minimizes the impact on eMBB performance. In addition, the proposed algorithm is compared with three reference algorithms with similar objectives and the performance is evaluated in terms of eMBB Spectral Efficiency, Sum throughput and Fairness level. The simulation results show that the proposed algorithm outperforms the above-mentioned reference algorithms in all evaluation metrics and showed its capability of elevating the performance of heuristic scheduling algorithms in the presence of uRLLC traffic. In the second part of this thesis, an optimal resource allocation scheme with guaranteed fairness is proposed in which, it can provide the desired level of fairness among eMBB users while maximizing their data rate. The results show how the fairness level and the number of uRLLC users affect the performance of the algorithm using different intensities. It also shows that the optimal allocation scheme provides better results in terms of eMBB sum-throughput while preserving the desired level of fairness.