Bayesian Online Learning for Energy-Aware Resource Orchestration in Virtualized RANs

Abstract

Radio Access Network Virtualization (vRAN) will spearhead the quest towards supple radio stacks that adapt to heterogeneous infrastructure: from energy-constrained platforms deploying cells-on-wheels (e.g., drones) or battery-powered cells to green edge clouds. We perform an in-depth experimental analysis of the energy consumption of virtualized Base Stations (vBSs) and render two conclusions: (i) characterizing performance and power consumption is intricate as it depends on human behavior such as network load or user mobility; and (ii) there are many control policies and some of them have non-linear and monotonic relations with power and throughput. Driven by our experimental insights, we argue that machine learning holds the key for vBS control. We formulate two problems and two algorithms: (i) BP-vRAN, which uses Bayesian online learning to balance performance and energy consumption, and (ii) SBP-vRAN, which augments our Bayesian optimization approach with safe controls that maximize performance while respecting hard power constraints. We show that our approaches are data-efficient and have provably performance, which is paramount for carriergrade vRANs. We demonstrate the convergence and flexibility of our approach and assess its performance using an experimental prototype.

Orchestrating Energy-Efficient vRANs: Bayesian Learning and Experimental Results

“…The source code of BP-vRAN and SBP-vRAN and the produced experimental datasets are publicly available, aspiring to facilitate the evaluation of other AI/ML solutions for vRAN orchestration. This paper extends our preliminary conference version [19] with the following contributions:…”

Network Intelligence for Virtualized RAN Orchestration: The DAEMON Approach

“…Our goal is to use O-RAN's control architecture to implement near-real-time configuration policies that are adaptive to system dynamics while satisfying hard energy constraints. Specifically, we consider the Safe Bayesian Optimization vRAN control algorithm (SBP-vRAN) recently introduced in [12], [14].…”

“…Other parametric methods, such as Reinforcement Learning relying on neural networks, need to be re-trained if the constraint changes, which substantially increases the required training data. The details of the learning model can be found in [12], [14].…”

“…In these settings, finding a suitable framework that maximizes the reuse of ML models across diverse NI solutions is of paramount importance: it paves the way for a structured approach to ML assimilation in networking systems, greatly simplifying future innovations in NI for beyond-5G and 6G networks. Specifically, we present mappings between the N-MAPE-K loop proposed as the unified NI representation within DAE-MON [10] and two sample NI algorithms developed in the project: vrAIn [11] and SBP-vRAN [12]. We show that the DAEMON architecture promotes the reuse of NI components, such as monitoring ones, among heterogeneous algorithms used inside each micro-domain.…”

“…The N-MAPE-K definition for the vrAIn[11], and SBP-vRAN[12] algorithms. SNR measurements, traffic demands: as Buffer State Reports (BSRs) from the terminals.…”