Assessing Power Allocation Efficiency for RAN in Cloud-based Systems for 5G Networks

João Pedro Albuquerque, Glauco Estácio Gonçalves, Aldebaro Klautau

DOI: 10.14209/sbrt.2025.1571157263

Evento: XLIII Simpósio Brasileiro de Telecomunicações e Processamento de Sinais (SBrT2025)

Keywords: 5G Radio Access Network Power Allocation Cloud

Abstract

Deploying edge and cloud computing architectures in 5G networks offers significant advantages, including reduced latency, lower core network traffic, and distributed processing capabilities. However, relocating latency-sensitive Radio Resource Management (RRM) functions, such as power allocation, to edge or cloud nodes may degrade the channel capacity and performance of User Equipments (UEs). This paper evaluates the impact of allocating a power allocation function at different network levels-Radio Access Network (RAN), Mobile Edge Computing (MEC), and cloud-using the ns-3 simulator with the 5G-Lena module. We implemented a simple, memoryless power allocation algorithm to assess channel capacity variations under varying cloud-distance latency conditions. We analyzed key performance indicators, including the Round-Trip Time (RTT) of control packets and channel capacity over time, to investigate the impact of latency on power allocation efficiency. The results reveal that allocating the power allocation function closer to the RAN achieves superior performance, with mean capacity values and variations meeting 3GPP standards. In contrast, placing the function at the MEC or cloud led to increased latency, insufficient capacity levels for some UEs, and more significant deviations from 3GPP requirements.

Download