The vision of intelligent networks capable of automatically configuring crucial parameters for tasks such as resource provisioning, anomaly detection or load balancing largely hinges upon efficient AI-based algorithms. Time series forecasting is a fundamental building block for network-oriented AI and current trends lean towards the systematic adoption of models based on deep learning approaches. In this paper, we pave the way for a different strategy for the design of predictors for mobile network environments, and we propose the Thresholded Exponential Smoothing (TES) framework, a hybrid Statistical Modeling and Deep Learning tool that allows for improving the performance of network Key Performance Indicator (KPI) forecasting. We adapt our framework to two state-of-the-art deep learning tools for time series forecasting, based on Recurrent Neural Networks and Transformer architectures. We experiment with TES by showcasing its superior support for three practical network management use cases, i.e. (i) anticipatory allocation of network resources, (ii) mobile traffic anomaly prediction, and (iii) mobile traffic load balancing. Our results, derived from traffic measurements collected in operational mobile networks, demonstrate that the TES framework can yield substantial performance gains over current state-of-the-art predictors in the applications considered.

Radio Access Network (RAN) virtualization, key for new-generation mobile networks, requires Hardware Accelerators (HAs) that swiftly process wireless signals from Base Stations (BSs) to meet stringent reliability targets. However, HAs are expensive and energy-hungry, which increases costs and has serious environmental implications. To address this problem, we gather data from our experimental platform and compare the performance and energy consumption of a HA (NVIDIA GPU V100) vs. a CPU (Intel Xeon Gold 6240R, 16 cores) for energy-friendly software processing. Based on the insights obtained from this data, we devise a strategy to offload workloads to HAs opportunistically to save energy while preserving reliability. This offloading strategy, however, needs to be configured in near-real-time for every BS sharing common computational resources. This renders a challenging multi-agent collaborative problem in which the number of involved agents (BSs) can be arbitrarily large and can change over time. Thus, we propose an efficient multi-agent contextual bandit algorithm called ECORAN1, which applies concepts from mean field theory to be fully scalable. Using a real platform and traces from a production mobile network, we show that ECORAN can provide up to 40% energy savings with respect to the approach used today by the industry.

RAN virtualization is revolutionizing the telco industry, enabling 5G Distributed Units to run using general-purpose platforms equipped with Hardware Accelerators (HAs). Recently, GPUs have been proposed as HAs, hinging on their unique capability to execute 5G PHY operations efficiently while also processing Machine Learning (ML) workloads. While this ambivalence makes GPUs attractive for cost-effective deployments, we experimentally demonstrate that multiplexing 5G and ML workloads in GPUs is in fact challenging, and that using conventional GPU-sharing methods can severely disrupt 5G operations. We then introduce YinYangRAN, an innovative O-RAN-compliant solution that supervises GPU-based HAs so as to ensure reliability in the 5G processing pipeline while maximizing the throughput of concurrent ML services. YinYangRAN performs GPU resource allocation decisions via a computationally-efficient approximate dynamic programming technique, which is informed by a neural network trained on real-world measurements. Using workloads collected in real RANs, we demonstrate that YinYangRAN can achieve over 50% higher 5G processing reliability than conventional GPU sharing models with minimal impact on co-located ML workloads. To our knowledge, this is the first work identifying and addressing the complex problem of HA management in emerging GPU-accelerated vRANs, and represents a promising step towards multiplexing PHY and ML workloads in mobile networks.

Open and virtualized Radio Access Networks (vRANs) are breeding a new market with unprecedented opportunities. However, carrier-grade vRANs today are expensive and energy-hungry, as they rely on hardware accelerators (HAs) that are dedicated to individual distributed units (DUs). In this paper, we argue that sharing pools of heterogeneous processors among DUs leads to more cost- and energy-efficient vRANs. We then design CloudRIC, a system that, powered by lightweight data-driven models, meets specific reliability targets while (i) coordinating access between DUs and heterogeneous computing infrastructure; and (ii) assisting DUs with compute-aware radio scheduling procedures. Experiments on a GPU-accelerated O-Cloud show that CloudRIC can achieve, respectively, 3x and 15x mean gains in energy- and cost-efficiency under real RAN workloads while ensuring 99.999% reliability even in dense scenarios.

Open and virtualized Radio Access Networks (vRANs) are breeding a new market with unprecedented opportunities. However, carrier-grade vRANs today are expensive and energy-hungry, as they rely on hardware accelerators (HAs) that are dedicated to individual distributed units (DUs). We demonstrate CloudRIC [17], a system that, powered by lightweight data-driven models, meets specific reliability targets while (i) coordinating access between DUs and heterogeneous computing infrastructure; and (ii) assisting DUs with compute-aware radio scheduling procedures. Using a user-friendly dashboard to control an experimental testbed remotely, we demonstrate that CloudRIC achieves comparable reliability performance to a DU-dedicated platform while offering up to 40x higher cost-efficiency and up to 6x higher energy efficiency when pooling resources for up to 70 DUs.

Forecasting is a task of ever increasing importance for the operation of mobile networks, where it supports anticipatory decisions by network intelligence and enables emerging zero-touch service and network management models. While current trends in forecasting for anticipatory networking lean towards the systematic adoption of models that are purely based on deep learning approaches, we pave the way for a different strategy to the design of predictors for mobile network environments. Specifically, following recent advances in time series prediction, we consider a hybrid approach that blends statistical modelling and machine learning by means of a joint training process of the two methods. By tailoring this mixed forecasting engine to the specific requirements of network traffic demands, we develop a Thresholded Exponential Smoothing and Recurrent Neural Network (TES-RNN) model. We experiment with TES-RNN in two practical network management use cases, i.e., (i) anticipatory allocation of network resources, and (ii) mobile traffic anomaly prediction. Results obtained with extensive traffic workloads collected in an operational mobile network show that TES-RNN can yield substantial performance gains over current state-of-the-art predictors in both applications considered.