The goal of this work is to propose an energy-efficient algorithm for dynamic computation offloading, in a multi-access edge computing scenario, where multiple mobile users compete for a common pool of radio and computational resources. We focus on delay-critical applications, incorporating an upper bound on the probability that the overall time required to send the data and process them exceeds a prescribed value. In a dynamic setting, the above constraint translates into preventing the sum of the communication and computation queues' lengths from exceeding a given value. Ultra-reliable low latency communications (URLLC) are also taken into account using finite blocklengths and reliability constraints. The proposed algorithm, based on stochastic optimization, strikes an optimal balance between the service delay and the energy spent at the mobile device, while guaranteeing a target out-of-service probability. Starting from a long-term average optimization problem, our algorithm is based on the solution of a convex problem in each time slot, which is provided with a very fast iterative strategy. Finally, we extend the approach to mobile devices having energy harvesting capabilities, typical of Internet of Things scenarios, thus devising an energy efficient dynamic offloading strategy that stabilizes the battery level of each device around a prescribed operating level.

We provide an overview of the most recent advancements and outcomes of the European 6G flagship project Hexa-X, on the topic of in-network Artificial Intelligence (AI) and Machine Learning (ML). We first provide a general introduction to the project and its ambitions in terms of use cases (UCs), key performance indicators (KPIs), and key value indicators (KVIs). Then, we identify the key specific challenges to realize, implement, and enable the native integration of AI and ML in 6G, both as a means for designing flexible, low-complexity, and reconfigurable networks ( learning to communicate ), and as an intrinsic in-network intelligence feature ( communicating to learn or, 6G as an efficient AI/ML platform). We present a high level description of down selected technical enablers and their implications on the Hexa-X identified UCs, KPIs and KVIs. Our solutions cover lower layer aspects, including channel estimation, transceiver design, power amplifier and distributed MIMO related challenges, and higher layer aspects, including AI/ML workload management and orchestration, as well as distributed AI. The latter entails Federated Learning (FL) and explainability as means for privacy preserving and trustworthy AI. To bridge the gap between the technical enablers and the 6G targets, some representative numerical results accompany the high level description. Finally, a brief discussion of the ongoing regulation activities related to AI, is presented, to close our vision towards an AI and ML-driven communication and computation co-design for 6G.

The aim of this paper is to propose a resource allocation strategy for dynamic training and inference of machine learning tasks at the edge of the wireless network, with the goal of exploring the trade-off between energy, delay and learning accuracy. The scenario of interest is composed of a set of devices sending a continuous flow of data to an edge server that extracts relevant information running online learning algorithms, within the emerging framework known as Edge Machine Learning (EML). Taking into account the limitations of the edge servers, with respect to a cloud, and the scarcity of resources of mobile devices, we focus on the efficient allocation of radio (e.g., data rate, quantization) and computation (e.g., CPU scheduling) resources, to strike the best trade-off between energy consumption and quality of the EML service, including service end-to-end (E2E) delay and accuracy of the learning task. To this aim, we propose two different dynamic strategies: (i) The first method aims to minimize the system energy consumption, under constraints on E2E service delay and accuracy; (ii) the second method aims to optimize the learning accuracy, while guaranteeing an E2E delay and a bounded average energy consumption. Then, we present a dynamic resource allocation framework for EML based on stochastic Lyapunov optimization. Our low-complexity algorithms do not require any prior knowledge on the statistics of wireless channels, data arrivals, and data probability distributions. Furthermore, our strategies can incorporate prior knowledge regarding the model underlying the observed data, or can work in a totally data-driven fashion. Several numerical results on synthetic and real data assess the performance of the proposed approach.

The aim of this paper is to propose a novel dynamic resource allocation strategy for energy-efficient adaptive federated learning at the wireless network edge, with latency and learning performance guarantees. We consider a set of devices collecting local data and uploading processed information to an edge server, which runs stochastic gradient-based algorithms to perform continuous learning and adaptation. Hinging on Lyapunov stochastic optimization tools, we dynamically optimize radio parameters (e.g., set of transmitting devices, transmit powers, bits, and rates) and computation resources (e.g., CPU cycles at devices and at server) in order to strike the best trade-off between power, latency, and performance of the federated learning task. The framework admits both a model-based implementation, where the learning performance metrics are available in closed-form, and a data-driven approach, which works with online estimates of the learning performance of interest. The method is then customized to the case of federated least mean squares (LMS) estimation, and federated training of deep convolutional neural networks. Numerical results illustrate the effectiveness of our strategy to perform energy-efficient, low-latency, adaptive federated learning at the wireless network edge.

We propose a novel strategy for energy-efficient dynamic computation offloading, in the context of edge-computing-aided beyond 5G networks. The goal is to minimize the energy consumption of the overall system, comprising multiple User Equipment (UE), an access point (AP), and an edge server (ES), under constraints on the end-to-end service delay and the packet error rate performance over the wireless interface. To reduce the energy consumption, we exploit low-power sleep operation modes for the users, the AP and the ES, shifting the edge computing paradigm from an always on to an always available architecture, capable of guaranteeing an on-demand target service quality with the minimum energy consumption. To this aim, we propose an online algorithm for dynamic and optimal orchestration of radio and computational resources called Discontinuous Computation Offloading (DisCO) . In such a framework, end-to-end delay constraints translate into constraints on overall queueing delays, including both the communication and the computation phases of the offloading service. DisCO hinges on Lyapunov stochastic optimization, does not require any prior knowledge on the statistics of the offloading traffic or the radio channels, and satisfies the long-term performance constraints imposed by the users. Several numerical results illustrate the advantages of the proposed method.

Private networks will play a key role in 5G and beyond to enable smart factories with the required better deployment, operation and flexible usage of available resource and infrastructure. 5G private networks will offer a lean and agile solution to effectively deploy and operate services with stringent and heterogeneous constraints in terms of reliability, latency, re-configurability and re-deployment of resources as well as issues related to governance and ownership of 5G components, and elements. In this paper, we present a novel approach to operator models, specifically targeting 5G and beyond private networks. We apply the proposed operator models to different network architecture options and to a selection of relevant use cases offering mixed private–public network operator governance and ownership. Moreover, several key enabling technologies have been identified for 5G private networks. Before the deployment, stakeholders should consider spectrum allocation and on-site channel measurements in order to fully understand the propagation characteristic of a given environment and to set up end-to-end system parameters. During the deployment, a monitoring tools will support to validate the deployment and to make sure that the end-to-end system meet the target KPI. Finally, some optimization can be made individually for service placement, network slicing and orchestration or jointly at radio access, multi-access edge computing or core network level.