Renewable Energy Assisted Function Splitting in Cloud Radio Access Networks

TL;DR

This paper explores integrating renewable energy sources into cloud radio access networks with function splitting to reduce operational costs and fronthaul bandwidth, proposing optimization models and heuristics for efficient decision-making.

Contribution

It introduces a novel renewable energy-assisted function splitting architecture in C-RAN, formulates an MILP model for cost reduction, and develops a heuristic for high user density scenarios.

Findings

Renewable energy integration reduces operational costs.

Function splitting with renewables improves cost-effectiveness.

Heuristic provides solutions for high user densities.

Abstract

Cloud-Radio Access Network (C-RAN) is a promising network architecture to reduce energy consumption and the increasing number of base station deployment costs in mobile networks. However, the necessity of enormous fronthaul bandwidth between a remote radio head and a baseband unit (BBU) calls for novel solutions. One of the solutions introduces the edge-cloud layer in addition to the centralized cloud (CC) to keep resources closer to the radio units (RUs). Then, split the BBU functions between the center cloud (CC) and edge clouds (ECs) to reduce the fronthaul bandwidth requirement and to relax the stringent end-to-end delay requirements. This paper expands this architecture by combining it with renewable energy sources in CC and ECs. We explain this novel system and formulate a mixed-integer linear programming (MILP) problem, which aims to reduce the operational expenditure of this system. Due to the NP-Hard property of this problem, we solve the smaller instances by using a MILP Solver and provide the results in this paper. Moreover, we propose a faster online heuristic to find solutions for high user densities. The results show that make splitting decisions by considering renewable energy provides more cost-effective solutions to mobile network operators (MNOs). Lastly, we provide an economic feasibility study for renewable energy sources in a CRAN architecture, which will encourage the MNOs to use these sources in this architecture.