5G New Radio Resource Allocation Optimization for Heterogeneous Services

TL;DR

This paper proposes a new resource allocation framework for 5G NR that efficiently manages heterogeneous services like URLLC and eMBB, balancing performance and computational complexity through optimization and heuristic methods.

Contribution

It introduces a novel formulation of 5G NR resource allocation considering service interplay and proposes a low-complexity heuristic inspired by bin packing for practical implementation.

Findings

The proposed methods meet QoS requirements probabilistically.

The heuristic balances resource efficiency and computational complexity.

Performance results validate the effectiveness of the approaches.

Abstract

5G new radio (NR) introduced flexible numerology to provide the necessary flexibility for multiplexing the communication of heterogeneous services on a shared channel. One of the fundamental challenges of 5G NR is to develop resource allocation schemes to efficiently exploit such flexibility to optimize resource allocation of ultra-reliable low-latency communications (URLLC) in coexistence with enhanced mobile broadband (eMBB) while ensuring their colliding performance requirements. To address this challenge, we present a new formulation of 5G NR resource allocation to accommodate eMBB and URLLC services, by considering their interplay. The objective of the formulated problem is to meet throughput quality of service (QoS) requirements probabilistically when the optimal global solution is infeasible. To this end, we express the problem as an integer linear program and consider two formulations with hard and soft URLLC throughput constraints. Furthermore, we propose a low-complexity two-step heuristic approach, inspired by an instance of bin packing optimization, to provide a trade-off between resource allocation efficiency and computational complexity. Finally, performance results are provided and demonstrate that the proposed approaches can provide low-complexity resource allocation solutions, in balance with the performance targets of the services.