Transmission and Latency-Aware Load Balancing for Fog Radio Access Networks

• Published in: 2018 IEEE Global Communications Conference (GLOBECOM) pp. 1 - 6
• Main Authors: Mukherjee, Mithun, Yejun Liu, Lloret, Jaime, Lei Guo, Matam, Rakesh, Aazam, Mohammad
• Format: Conference Proceeding
• Language: English; Japanese
• Published: IEEE 01-12-2018
• Subjects: Cloud computing; Delays; Edge computing; Interference; Radio access networks; Radio frequency; Task analysis
• ISSN: 2576-6813
• DOI: 10.1109/GLOCOM.2018.8647580

Fog computing-based radio access networks (F-RANs) aim to extend the computing and storage facilities of the centralized cloud radio access networks (C-RANs) to the network edge. Compared with the centralized baseband unit pool in the C-RAN, F-RAN reduces the burden on fronthaul. Thus, the F-RAN is foreseen as a viable solution towards ultra-low latency service provisioning. However, due to limited computing and storage facilities in fog computing-enabled access points (F-APs), some tasks that cannot be executed on the primary F-APs are transferred to other F-APs. In worst-case, the tasks are sent to the resource-enriched centralized cloud for processing. The transmission latency between F-APs, F-AP-to-end- user, and fronthaul latency strongly depends on interference power from the undesired network element as well as end-users. At the same time, the computational latency increases with the queuing delay. In this paper, we propose a load balancing scheme to address the tradeoff between transmission and computing latencies in F- RANs. Finally, the extensive simulation results show that the proposed scheme outperforms the greedy approach to meet the critical requirements, such as low-latency and minimal task offloading to the cloud in the F-RAN for the low-latency communications.