Energy-Efficient Orchestration of Metro-Scale 5G Radio Access Networks

Abstract

RAN energy consumption is a major OPEX source for mobile telecom operators, and 5G is expected to increase these costs by several folds. Moreover, paradigm-shifting aspects of the 5G RAN architecture like RAN disaggregation, virtualization and cloudification introduce new traffic-dependent resource management decisions that make the problem of energy-efficient 5G RAN orchestration harder. To address such a challenge, we present a first comprehensive virtualized RAN (vRAN) system model aligned with 5G RAN specifications, which embeds realistic and dynamic models for computational load and energy consumption costs. We then formulate the vRAN energy consumption optimization as an integer quadratic programming problem, whose NP-hard nature leads us to develop GreenRAN, a novel, computationally efficient and distributed solution that leverages Lagrangian decomposition and simulated annealing. Evaluations with real-world mobile traffic data for a large metropolitan area are another novel aspect of this work, and show that our approach yields energy efficiency gains up to 25% and 42%, over state-of-the-art and baseline traditional RAN approaches, respectively.

Virtualized Radio Access Networks: Energy Models, Challenges and Opportunities

“…The isolation between virtual machines (VMs) or containers can take place at the function or RRH boundary. For example, [41] considers a BBU that is split into DU and CU components, with each edge cloud being responsible for the processing of virtual network functions (VNFs) for a specific set of mutually exclusive BSs. [41] presents a computational load model (Equation 10) that is given by the sum of the processing load of all VNFs, depending on the functional split.…”

“…For example, [41] considers a BBU that is split into DU and CU components, with each edge cloud being responsible for the processing of virtual network functions (VNFs) for a specific set of mutually exclusive BSs. [41] presents a computational load model (Equation 10) that is given by the sum of the processing load of all VNFs, depending on the functional split. In addition, the computation load of an edge or cloud machine is given as the sum of the computation load of their associated VMs.…”

Slice Aware Baseband Function Placement in 5G RAN Using Functional and Traffic Split

“…The functional splits are used to propose a three-tier architecture for 5G RAN We consider a hybrid cloud architecture as our system model (shown in Fig. 1) along with the three-tier architecture for RAN, which conforms to the 5G RAN deployment scenario [28], [29]. The multiple RUs in the network denote the cell sites.…”

“…These layers are placed in the DU or CU based on the functional split resulting in four different functional splits. We consider RRC and PDCP together due to less processing requirement of RRC layer [28]. We assign split numbers 0-3, with 0 being the lowest split and 3 being the highest.…”

“…The edge clouds are connected to the regional cloud through the midhaul network. The midhaul network consists of multiple paths which have delays in the range of 2-30 ms and aggregated path capacities ranging from 2-28 Gbps [28], [33]. There are three slices in each RU cluster for eMBB, URLLC, and mMTC services, respectively [6].…”