This paper explores the role of multivariate probabilistic forecasting in improving O-RAN operations, focusing on network sustainability aspects. A comprehensive analysis of its potential benefits and challenges, as well as its integration into the O-RAN architecture are described. The paper first presents an overview of the O-RAN architecture and components, followed by an examination of power consumption models relevant to O-RAN deployments and the challenges associated with traditional deterministic models in resource allocation. We then examine the performance of several state-of-the-art probabilistic multivariate forecasting techniques namely, Gaussian Process Vector Autoregression (GPVAR), Temporal Fusion Transformer (TFT) and non-probabilistic multivariate technique namely, Multivariate Long-Short Term Memory (LSTM) and explain their implementation details and provide their evaluations. The simulation results show the effectiveness of these techniques in predicting Physical Resource Block (PRB) utilization and optimizing resource allocation. In particular, significant energy savings – around 20-30%– are achieved, depending on the percentile of the used probabilistic forecasting techniques. The benefits of considering probabilistic forecasting techniques compared to multi-variate LSTM are also analyzed. Our results emphasize the potential of probabilistic forecasting to improve energy efficiency and sustainability in O-RAN operations.

In this paper, we present an innovative federated learning (FL) approach that utilizes Kolmogorov-Arnold Networks (KANs) for classification tasks. By utilizing the adaptive activation capabilities of KANs in a federated framework, we aim to improve classification capabilities while preserving privacy. The study evaluates the performance of federated KANs (F-KANs) compared to traditional federated Multi-Layer Perceptrons (F-MLPs) on classification task. The results show that the F-KANs model significantly outperforms the F-MLP model in terms of accuracy, precision, recall, F1 score and stability, and achieves better performance, paving the way for more efficient and privacy-preserving predictive analytics.

The inherent virtualization characteristics of 5G, 6G, and subsequent generations (xG) facilitate the modularity of components and their interfaces, leading to an increase in the number of stakeholders and necessitating more complex administrative relationships. In this demonstration, we present an open-source Decentralized Application (DApp) integrated with smart contracts deployed on a live Polygon testnet to support trustworthy Federated Learning (FL). FL enables clients to conduct local training and updates, while a central aggregation server processes these inputs for global FL training. We introduce a blockchain-based reputation system within such FL model to facilitate collaborative training of Machine Learning (ML) models. The deployed smart contracts on the Layer 2 Polygon blockchain calculate and store reputation scores on-chain for each FL client based on their contribution. This demonstration presents a visualization of the blockchain network parameters during the execution of blockchain-enabled smart contracts. The DApp offers an interactive interface for client registration, performance parameter submission, and reputation score calculation. We use Alchemy's dashboard for real-time monitoring of blockchain transactions and smart contract interactions. The open-source implementation is publicly available11https://github.com/farhanajaved/bc-fl-demo-l2 and the video of the demo is available at22https://youtu.be/B9J1gxjoKdM.

The virtualization and softwarization of 5G/6G mobile networks have enabled the deployment and orchestration of cloud-native network and application functions. The deployment of these functions is crucial, as the placement of data plane elements (i.e., User Plane Function (UPF)) and vertical services can significantly impact the overall user latency. However, in multi-slice edge scenarios, characterized by users with distinct levels of criticality, the problem of UPF and application placement is becoming increasingly complex due to i) the various costs involved and ii) the limited computational resources at the edge. In this paper, the problem of joint UPF and application placement for a multi-slice user scenario is studied, taking into account multiple cost components that influence the placement decision, including service migration, traffic forwarding, server activation and processing costs. To tackle this problem, we introduce a Joint UPF and Application Reinforcement Learning-based (JUAP-RL) algorithm, which decides the UPF and application deployment location and coordinates the placement stages. Extensive experiments have shown that JUAP-RL demonstrates up to 17% gain in terms of user acceptance ratio and up to 23.4% reduction in provisioning cost compared to baseline schemes.

Resource provisioning plays a crucial role in effective resource management. As we move into the 6G era, technologies such as Open Radio Access Network (O-RAN) offer the opportunity to develop intelligent and interoperable cutting-edge solutions for qualitative management of the latest communication system. Previous works have mostly used single-point forecasts like Long-Short Term Memory (LSTM) for predicting resource requirements, which presents decision-makers with the problem of making informed decisions about resource allocation. On the other hand, probability-based forecasting techniques such as DeepAR, Transformer and Simple-Feed-Forward (SFF) offer new dimensions to the predictions by quantifying their uncertainties. This work shows the comprehensive comparison of single-point and probabilistic estimators and evaluates their effectiveness in predicting the actual number of Physical Resource Blocks (PRBs) needed in the context of O-RAN, especially for multi-tenant use cases. The results show the superiority of the probabilistic model in terms of various evaluation metrics. DeepAR achieves the highest accuracy, outperforming single-point and other probabilistic estimators. Based on these findings, a novel approach named Dynamic Percentile Adjustment Approach (DYNp) algorithm is proposed, which utilizes probabilistic forecasting for adaptive resource allocation. After extensive analysis, the numerical results show that the DYNp algorithm for DeepAR predictions reduces the Service Level Agreement (SLA) violation to 8% and the over-provisioning to 0.509 by dynamic percentile adaption. DYNp approach ensures that resources are allocated by efficiently handling over-and under-provisioning, making it suitable for real-time scenarios in O-RAN environments.

This paper deals with the problem of energy consumption minimization in Open RAN cell-free (CF) massive Multiple-Input Multiple-Output (mMIMO) systems under minimum per-user signal-to-noise-plus-interference ratio (SINR) constraints. Considering that several access points (APs) are deployed with multiple antennas, and they jointly serve multiple users on the same time-frequency resources, we design the precoding vectors that minimize the system power consumption, while preserving a minimum SINR for each user. We use a simple, yet representative, power consumption model, which consists of a fixed term that models the power consumption due to activation of the AP and a variable one that depends on the transmitted power. The mentioned problem boils down to a binary-constrained quadratic optimization problem, which is strongly non-convex. In order to solve this problem, we resort to a novel approach, which is based on the penalized convex-concave procedure. The proposed approach can be implemented in an O-RAN cell-free mMIMO system as an xApp in the near-real time RIC (RAN intelligent Controller). Numerical results show the potential of this approach for dealing with joint precoding optimization and AP selection.

In this paper, we propose an innovative integration of Quantum Key Distribution (QKD) and Blockchain-based Self-Sovereign Identity (SSI) within the Open RAN (O-RAN) framework for 6G networks to address the critical need for enhanced security and robust identity management. We first present a general architecture that takes a multi-layered approach and is carefully designed to leverage the different capabilities of quantum security and blockchain technology. The architecture ensures seamless and secure operation across different layers of the O-RAN, focusing on the Distributed Identity Management (DIM) and Management & Orchestration layers, and explains the interactions between these layers to improve the security and operational efficiency of the network. We also investigate detailed case studies and applications that demonstrate the practicality and transformative potential of integrating QKD-secured blockchain identity management systems in real-world 6G scenarios. We also address the inherent challenges and limitations of such integration and propose viable solutions to overcome them. Finally, we provide insights into future research and implementation directions and highlight the critical role of quantum-secured blockchain systems in the evolution of telecommunication networks toward a more secure, decentralized, and user-centric paradigm.

The transition to sustainable Open Radio Access Network (O-RAN) architectures brings new challenges for resource management, especially in predicting the utilization of Physical Resource Block (PRB)s. In this paper, we propose a novel approach to characterize the PRB load using probabilistic forecasting techniques. First, we provide background information on the $O-R A N$ architecture and components and emphasize the importance of energy/power consumption models for sustainable implementations. The problem statement highlights the need for accurate PRB load prediction to optimize resource allocation and power efficiency. We then investigate probabilistic forecasting techniques, including Simple-Feed-Forward (SFF), DeepAR, and Transformers, and discuss their likelihood model assumptions. The simulation results show that DeepAR estimators predict the PRBs with less uncertainty and effectively capture the temporal dependencies in the dataset compared to SFF- and Transformer-based models, leading to power savings. Different percentile selections can also increase power savings, but at the cost of over-/under provisioning. At the same time, the performance of the Long-Short Term Memory (LSTM) is shown to be inferior to the probabilistic estimators with respect to all error metrics. Finally, we outline the importance of probabilistic, prediction-based characterization for sustainable O-RAN implementations and highlight avenues for future research.

The H2020 Distributed management of Network Slices in beyond 5G(MonB5G) project aims to provide zero-touch management and orchestration to support network slicing at scale to reduce the management burden on mobile operators by leveraging distribution of operations along with advanced data-driven Artificial Intelligence (AI)-based mechanisms. However, while this approach shows promise and large companies with abundant data and ML expertise are developing powerful MLdriven services, a critical aspect that remains to be analyzed is its business case. The vast majority of potentially valuable ML services, such as predictive maintenance, Quality of Service (QoS) optimization, network security enhancements, remain stuck at the idea or prototype stage. This paper delves into an analysis of how the MonB5G solutions in particular the tuples (Monitoring System (MS), Analytics Engine (AE), Decision Engine (DE) and Actuator (ACT) could be applied within the network management and orchestration market while investigating various business models and value chains. Numerical results based on experimental data have also been performed to evaluate the OpEX (Operational Expenditure) benefits associated with different network management techniques, for centralized and distributed systems.

Self-Sovereign Identity (SSI) has emerged lately as an identity and access management framework that is based on Distributed Ledger Technology (DLT) and allows users to control their own data. Federate Learning (FL), on the other hand, provides a framework to update Machine Learning (ML) models without relying on explicit data exchange between the users. This paper investigates identity management and authentication for vehicle users, which are participating into FL. We propose a new approach to SSI, that is alternative to the conventional blockchain-based SSI, specifically for use in vehicular networks, which focuses on maintaining confidentiality, authenticity, and integrity of vehicle users' identities and data exchanged between the users and the aggregation server during the execution of the FL process. We also provide experimental results for distributed identity management (DIM) operations, which show that the performance of credential operations in the implemented system is generally efficient and the average times are within reasonable limits. However, there is a slight increase in presentation time, offer time, connection establishment time, and credential revocation time as the number of requests increases, indicating a slight degradation in performance for these operations.

This paper demonstrates an implementation of Federated Learning (FL) for streaming applications using cloud-native technology. Compared to a centralized management, by adopting a decentralized approach, the FL method improves convergence time, reduces communication overhead, and increases network energy efficiency. The cloud-native FL architecture presented comprises three sites, each with its own Kubernetes (K8s) cluster. The edge sites run FL Analytical Engines (AEs)/clients for local training and updates, and the central site runs the aggregation server for FL training. Some other relevant workloads deployed at the clusters are the video streaming server, the orchestrator, and monitoring components. As for the RAN, we showcase a multi-gNB setup from which we obtain monitoring data via custom sampling functions. Following the description of the testbed infrastructure and setup, this demonstration presents the real-time visualization of network parameters during FL training, and the enhancement of video streaming through proactive Central Processing Unit (CPU) scaling, made possible by the resource forecasting.

A key enabler for both scalability and sustainability in beyond 5G (B5G) network slicing consists on minimizing the exchange of raw monitoring data across different domains. This is achieved by bringing the analysis functions closer to the data collection points. To this end, we introduce in this paperstatistical federated learning(SFL) provisioning models that can learn over a live network non independent identically distributed (non-IID) datasets in an offline fashion while respecting slice-level service level agreement (SLA) long-term statistical constraints. Specifically, we consider three resource SLA metrics, namely,cumulative distribution function(CDF),$Q$-th percentileandmaximum/minimum bounds. These metrics are dataset-dependent and non-convex non-differentiable and, to sidestep the inaccuracy of settling only for surrogates, we propose a novel formulation that jointly considers the statistical objective and constraints as well as their smooth approximation using theproxy-Lagrangianframework, which we solve via a non-zero sum two-player game strategy. Numerical results on various slice-level resources show that SFL enables SLA enforcement while significantly reducing the overhead compared to both state-of-the-art FedAvg and centralized constrained deep learning schemes. Finally, we provide an analysis for the lower bound of the so-calledreliable convergence probabilityin the SFL setup.

Artificial intelligence (AI)-driven zero-touch massive network slicing is envisioned to be a disruptive technology in beyond 5G (B5G)/6G, where tenancy would be extended to the final consumer in the form of advanced digital use-cases. In this paper, we propose a novel model-free deep reinforcement learning (DRL) framework, called collaborative statistical Actor-Critic (CS-AC) that enables a scalable and farsighted slice performance management in a 6G-like RAN scenario that is built upon mobile edge computing (MEC) and massive multiple-input multiple-output (mMIMO). In this intent, the proposed CS-AC targets the optimization of the latency cost under a long-term statistical service-level agreement (SLA). In particular, we consider the Q-th delay percentile SLA metric and enforce some slice-specific preset constraints on it. Moreover, to implement distributed learners, we propose a developed variant of soft Actor-Critic (SAC) with less hyperparameter sensitivity. Finally, we present numerical results to showcase the gain of the adopted approach on our built OpenAI-based network slicing environment and verify the performance in terms of latency, SLA Q-th percentile, and time efficiency. To the best of our knowledge, this is the first work that studies the feasibility of an AI-driven approach for massive network slicing under statistical SLA.

In this paper, we address the concept of dynamic resource allocation for radio access network (RAN) slicing in beyond 5G (B5G) systems under service-level agreement (SLA). Using live network distributed key performance indicators (KPIs) mini-datasets, we introduce a new class of federated learning models that can capture the long-term cumulative distribution function (CDF) statistic—is usually used to define SLA—and enforce some preset constraints on it. Given that the CDF is also dataset-dependent and non-convex non-differentiable, we formulate the corresponding local optimization task using the proxy-Lagrangian framework and solve it via a non-zero sum two-player game strategy. Numerical results show that the proposed decentralized resource allocation approach enables SLA enforcement and significantly reduces the SLA violation rate for various slice-level KPIs.

Next generation satellite payload technology is expected to provide digital processing capabilities. This will pave the way to regard satellites as flying base stations. However, the initial access procedure must be improved. In this work we focus on non-geostationary earth orbit (NGEO) satellite networks. In this scenario, the random access preamble signal and the detection must be robust to large carrier frequency offsets (CFOs). Towards this end, we investigate the adoption of the pruned discrete Fourier transform spread filter bank multicarrier waveform. The proposed design is suitable for the access scheme of forthcoming 5G-based NGEO satellite communications. The reason is twofold. First, it improves the spectral confinement with respect to the standard single-carrier frequency-division multiplexing (SC-FDM) waveform. Second, it achieves a high level of commonality with 5G new radio, by keeping unchanged the subcarrier spacing, the slot duration and the preamble sequence. Remarkably, the new design allows the straightforward application of non-coherent post detection integration (NCPDI) techniques, which divide the correlation in blocks. Numerical results show that the proposed solution reduces out-of-band emissions and the missed detection probability in presence of CFO, with respect to the conventional approach based on SC-FDM and preamble detection with full-length correlation.

Owing to the rapidly increasing traffic demands on satellite connectivity, the current exclusive frequency allocation is becoming obsolete. Instead, aggressive frequency reuse and interference mitigation techniques are promising ideas that both industry and academia are investigating. This paper proposes an optimization precoding technique for dealing with the multibeam interference due to the aggressive frequency reuse. In contrast to general multiuser multiple input multiple output (MIMO) schemes, multibeam satellite precoding techniques call for frame-by-frame quadratically constrained quadratic optimization of a large number of variables. We focus on the multigroup multicast beamforming optimization problem, and we propose to adopt a consensus-based alternating direction method of multipliers (C-ADMM) approach, in order to mitigate complexity. The proposed C-ADMM approach is shown to exhibit comparable optimization performance at considerably lower complexity relative to the prior state-of-art for the formulation considered.

This paper deals with the multiple relay selection problem in two-hop wireless cooperative networks with individual power constraints at the relays. In particular, it addresses the problem of selecting the best subset of K cooperative nodes and their corresponding beamforming weights so that the signal-to-noise ratio (SNR) is maximized at the destination. This problem is computationally demanding and requires an exhaustive search over all the possible combinations. In order to reduce the complexity, a new suboptimal method is proposed. This technique exhibits a near-optimal performance with a computational burden that is far less than the one needed in the combinatorial search. The proposed method is based on the use of the l1-norm squared and the Charnes-Cooper transformation and naturally leads to a semidefinite programming relaxation with an affordable computational cost. Contrary to other approaches in the literature, the technique exposed herein is based on the knowledge of the second-order statistics of the channels and the relays are not limited to cooperate with full power.

Positioning estimation based on time of arrival (TOA) estimation needs high accurate first arrival detector. High-resolution TOA estimators based on minimum variance (MV) and normalized minimum variance (NMV) provide an accurate estimation of the first arrival even in high multipath environments at the expenses of a high computational cost. The aim of this paper is to reduce the computational burden of high resolution TOA estimators. First, reduced complexity MV and NMV implementations based on the FFT are presented. Next, polynomial root versions of both estimators are proposed yielding an improvement in positioning accuracy. Finally, a near TOA maximum likelihood (ML) estimator is proposed providing a good trade off between complexity and accuracy.