Mixture-of-Experts (MoE) architectures are increasingly deployed across 6G edge–cloud networks, where sparse activation reduces the computational footprint of each inference to only a fraction of the full expert set. However, MoE inference in edge-cloud networks creates a tension between KV state locality and elastic expert dispatch. KV state relocation incurs substantial transfer overhead, while expert computation benefits from spreading across the network to exploit available capacity. This tension is amplified in multi-turn inference, where each turn extends the KV state that must persist across the dialogue. To this end, we present StateFlow, a distributed inference policy that decouples persistent KV state from transient sparse computation. StateFlow pins KV state at a sticky serving site for cross-turn reuse and jointly optimizes expert dispatch and aggregation placement across the network. We further implement a real-world testbed with kernel-level network emulation and experimental results show that StateFlow sustains more than 2× higher stable dialogue concurrency than the distributed baseline solutions and reduces turn-level p95 latency under multi-turn inference by 53.0%.

In this paper we propose a technical framework for applying formally verifiable AI (FVAI) methods to certify AI models that support 6G network functions. The certifications state that the input/output behavior of an AI model conforms to a specification or identifies potential issues with the model. This paper contributes to the broad spectrum of methods necessary to ensure that deep learning models, known to have instability issues, are sufficiently trustworthy for supporting functions in critical infrastructure like 6G networks. The proposed framework gives a unified way of representing network functions supported by AI models, certifications, construction of those certifications from model behaviour, and incorporating technical tools for bounding AI model input/output behaviour. We also illustrate the use of this framework in verifying AI models in two 6G use cases on optimisations in virtual network functions and radio access network resources. Furthermore, we discuss how this framework can be extended in the future.

Artificial Intelligence (AI) is expected to play a key role in 6G networks including optimising system management, operation, and evolution. This requires systematic lifecycle management of AI models, ensuring their impact on services and stakeholders is continuously monitored. While current 6G initiatives introduce AI, they often fall short in addressing end-to-end intelligence and crucial aspects like trust, transparency, privacy, and verifiability. Trustworthy AI is vital, especially for critical infrastructures like 6G. This paper introduces the REASON approach for holistically addressing AI's native integration and trustworthiness in future 6G networks. The approach comprises AI Orchestration (AIO) for model lifecycle management, Cognition (COG) for performance evaluation and explanation, and AI Monitoring (AIM) for tracking and feedback. Digital Twin (DT) technology is leveraged to facilitate real-time monitoring and scenario testing, which are essential for AIO, COG, and AIM. We demonstrate this approach through an AI-enabled xAPP use case, leveraging a DT platform to validate, explain, and deploy trustworthy AI models.

Future networks (including 6G) are poised to accelerate the realisation of Internet of Everything. The latter will imply a high demand for computational resources to support new services. Mobile Edge Computing (MEC) is a promising solution that enables offloading computation-intensive tasks to nearby edge servers from the end-user devices, thereby reducing latency and energy consumption. Nevertheless, relying solely on a single MEC server for task offloading can lead to uneven resource utilisation and suboptimal performance in complex scenarios. Additionally, traditional task offloading strategies specialise in centralised policy decisions, which unavoidably entails extreme transmission latency and reach computational bottleneck. To address these gaps, we propose a latency-efficient and energy-efficient Cooperative Task Offloading framework with Transformer-driven Prediction (CTO-TP), leveraging asynchronous multi-agent deep reinforcement learning to address these challenges. This approach fosters edge-edge cooperation and decreases the synchronous waiting time by performing asynchronous training, optimising task offloading, and resource allocation across distributed networks. The performance evaluation demonstrates that the proposed CTO-TP algorithm reduces up to 80% overall system latency and 87% energy consumption compared to the baseline schemes.

The emergence of beyond 5G and 6G networks is set to revolutionise telecommunications, addressing the demands of emerging applications through advanced capabilities. At the core of this transformation lies next-generation intelligent service orchestration, which is essential for meeting future Key Performance Indicators (KPIs) and Key Value Indicators (KVIs) such as ultra-low latency, efficient power consumption and resource utilization. These capabilities require multi-objective, seamless end-to-end service delivery across complex, distributed environments. Achieving such delivery requires scalable and modular system design approaches that support dynamic service composition and adaptability. Cloud-native technologies, underpinned by microservices architectures, plays a pivotal role, but also will introduce challenges in orchestrating resources efficiently across heterogeneous domains. To address these challenges, this paper proposes a solution, Federated Intelligent multi-objective Service function chain Orchestration (FISO) that integrates multi-objective federated profiling to preserve privacy while ensuring efficient end-to-end service delivery. FISO integrates Federated Learning (FL) and Reinforcement Learning (RL). FL is used to collaboratively learn from distributed edge profiling clients without sharing raw data, while RL dynamically guides optimal decision making for resource allocation and Service Function Chain (SFC) placement based on feedback from the federated models. FISO predicts optimal computing and network resources for SFCs, enabling the selection of appropriate edge locations, efficient resource allocation, placement of SFCs, and lifecycle management. Experimental results demonstrated on a pragmatic testbed validate the effectiveness of FISO in efficiently placing requested SFCs within an administrative domain with multiple edge/cloud nodes, predicting optimal CPU, memory, and link capacity resources, and minimizing end-to-end latency and energy consumption.

Future Open Networks (FONs) are envisioned as large-scale, decentralised, and data-driven systems designed with open interfaces and interoperable standards to deliver diverse services and applications. The management and orchestration of these networks will utilise intelligent processing. This paper introduces a novel Cognitive Cross-Domain Orchestration (CDO) architectural framework for FONs, structured as a quad-view architecture. A background setting for the architecture is provided by a survey of relevant projects, followed by a requirements analysis for FON orchestration and management. The proposed CDO architectural framework consists of four distinct views: the Organizational View identifies stakeholders and business entities; the Functional View outlines the logical design; the Cognitive & AI/ML View connects logical functions with supportive cognitive functions for network optimization; and the System-level View links these functions to the necessary physical and intangible assets. This segregation enables each view to provide a focused analysis of the roles, interactions, and properties, thereby promoting a coherent design. The paper also describes a plausible partial proof-of-concept implementation of the CDO System-level architecture. Within the context of this CDO implementation, stakeholder participation in the provision of an end-to-end service delivery orchestrated across multiple network domains is described.

Cryptographic network protocols play a crucial role in enabling secure data exchange over insecure media in modern network environments. However, even minor vulnerabilities can make protocols an easy target for cyber attackers. Therefore, it is essential to investigate the threats and vulnerabilities stemming from the cryptographic network protocols. Furthermore, it is necessary to comprehensively investigate the weaknesses of network protocols that use cryptographic primitives to inform users and developers about potential attack points. This comprehensive survey examines the relationship between encryption schemes and network protocols and presents an in-depth review of associated threats and vulnerabilities. Given that most cryptographic protocols operate in the Transport and Application layers of the Transmission Control Protocol/Internet Protocol (TCP/IP) protocol stack, our investigation primarily centers around encryption algorithms used by representative and notable cryptographic network protocols such as Transport Layer Security (TLS) and Secure Shell (SSH). Furthermore, we delve into the attackers’ methods to exploit the already identified and existing vulnerabilities, seeking to understand the mechanisms employed to compromise these protocols. Through this survey, we aim to provide the readership with an in-depth understanding of the existing and new vulnerabilities associated with modern cryptographic protocols and provide valuable insights into securing them effectively. We also discuss the existing challenges and future research directions in this domain.

Multiplayer computer games can be divided into two architectural groups: client-server and Peer-to-Peer (P2P). While the peer-to-peer approach is very promising due to its independence of expensive game servers, a number of issues, such as state synchronization in an un-trusted environment, arise. Blockchain networks are a special case of a P2P network, and they have the potential to efficiently tackle some issues with P2P-based multiplayer games. However, blockchain development is a vast and complicated area, foreign to most game developers. In this paper, we present a framework to connect an existing game client to a blockchain network, friendly towards game developers. More precisely, we develop a middleware solution and a data model therein, to store the game state on a chain and communicate it to the game client.

Federated Learning (FL) enables collaborative model training across edge devices while preserving data locally. Deploying FL faces challenges due to device heterogeneity. Using cloud technologies like Kubernetes (K8s) can offer computational elasticity, yet may compromise FL privacy principles. K8s can jeopardise FL privacy by potentially allowing malicious FL clients to access other resources given its flat networking approach. This paper introduces the privacy-preserving K8s operator kubeFlower. It addresses privacy risks via isolation-by-design and differential privacy for data management. Isolation ensures secure resource sharing, while differential privacy safeguards individual data privacy. We introduce the Privacy Preserving Persistent Volume Claimer (P3-VC), which adds noise to data while managing a privacy budget. kubeFlower simplifies FL system management in K8s while ensuring privacy. We tested our approach on a network testbed composed of different geo-located cloud and edge nodes where FL clients are deployed. Our results demonstrate the approach’s efficacy in preserving privacy in K8s-based FL for cloud–edge environments.

Federated Learning (FL) is vulnerable to various attacks including poisoning and inference. However, the existing offensive security evaluation of FL assumes that the attackers know data distribution. In this paper, we present a novel attack where FL participants carry out inference and privacy abuse attacks against the FL by leveraging Generating Adversarial Networks (GANs). The attacker (impersonating a benign participant) uses GAN to generate a similar dataset to other participants, and then covertly poisons the data. We demonstrated the attack successfully and tested it on two datasets, the IoT network traffic dataset and MNIST. The results reveal that for FL to be successfully used in IoT applications, protection against such attacks is critically essential.

Federated Learning (FL) is a privacy-enhancing technology for distributed ML. By training models locally and aggregating updates - a federation learns together, while bypassing centralised data collection. FL is increasingly popular in healthcare, finance and personal computing. However, it inherits fairness challenges from classical ML and introduces new ones, resulting from differences in data quality, client participation, communication constraints, aggregation methods and underlying hardware. Fairness remains an unresolved issue in FL and the community has identified an absence of succinct definitions and metrics to quantify fairness; to address this, we propose Federated Fairness Analytics - a methodology for measuring fairness. Our definition of fairness comprises four notions with novel, corresponding metrics. They are symptomatically defined and leverage techniques originating from XAI, cooperative game-theory and networking engineering. We tested a range of experimental settings, varying the FL approach, ML task and data settings. The results show that statistical heterogeneity and client participation affect fairness and fairness conscious approaches such as Ditto and q-FedAvg marginally improve fairness-performance trade-offs. Using our techniques, FL practitioners can uncover previously unobtainable insights into their system's fairness, at differing levels of granularity in order to address fairness challenges in FL. We have open-sourced our work at: https://github.com/oscardilley/federated-fairness.

Cyber-physical systems (CPSs) are being increasingly adopted for industrial applications, yet they involve a dynamic threat landscape that requires CPSs to adapt to emerging threats during their operation. Recently, digital twin (DT) technology (which refers to a virtual representation of a product, process, or environment) has emerged as a suitable candidate to address the security challenges faced by dynamic CPSs. DT has the capability of strengthening the security of CPSs by continuously mapping the physical to twin counterparts to detect inconsistencies. The existing DT-based security solutions are constrained by untrustworthy data dissemination as well as limited data sharing among the involved stakeholders, which, in turn, limit the ability of DTs to run accurate simulations or make valid decisions. To address these challenges, this paper proposes a modular framework called TR usted and I ntelligent cyber-PhysicaL systEm (TRIPLE), that leverages blockchain, DTs, and threat intelligence (TI) to secure CPSs. The blockchain-based DT components in the framework provide data integrity, traceability, and availability for trusted DTs. Furthermore, to accurately and comprehensively model system states, the framework envisions fusing process knowledge for modeling DTs from system specification- based and learning-based information and other sources, including infrastructure-as-code (IaC) and knowledge base (KB). The framework also integrates TI for future-proofing against emerging threats, such that threats can be detected either reactively by mapping the behavior of physical and virtual spaces or proactively by TI and threat hunting. We demonstrate the viability of the framework through a proof of concept. Finally, we formally verify the TRIPLE framework to demonstrate its correctness and effectiveness in enhancing CPS security.

The emergence of 6G networks promises to revolutionise telecommunications by incorporating end-to-end intelligence for improved performance, driven by new radio and other access technologies. On the other hand, optical technologies are crucial for enabling long-distance integration among various domains and network technologies. This integration is vital for future networks to meet Key Performance Indicators (KPIs) like bitrate and latency, and Key Value Indicators (KVIs) such as reliability and privacy. The envisioned end-to-end intelligence can drive optical networks to consistently provide high-quality connectivity and handle dynamic computing and network demands in 6G. This paper explores end-to-end intelligence within 6G networks, emphasising its potential to optimise service delivery. We envision a 6G-ready architecture for delivering end-to-end services across multiple domains interconnected by optical technologies. This architecture uses Artificial Intelligence and Machine Learning to support lifecycle management of infrastructures and network services. It includes an Intelligent Optical Cross-Domain Orchestrator and Optical Controllers at the Transport and Metro domains, as well as Intra-Domain and Cross-Domain Service Orchestration, efficiently allocating optimum resources to services ensuring robust and seamless performance in future networks.

Group Key Exchange (GKE) is an important tool to develop secure multi-user applications such as group text messages, ad-hoc networks, and so on.Most of the currently deployed GKE schemes are synchronous, i.e., they require all the participants to be online during their execution.However, with more battery-powered devices being used in such applications, the synchronicity requirement is challenging to fulfill.To fill the gaps, asynchronous GKE schemes have been introduced in the literature.Nevertheless, the currently available asynchronous and synchronous GKE schemes rely on Trusted Third Parties (TTPs) for key establishment and management.To this end, reliance on TTPs is a serious shortcoming since TTPs are well known to be the single point of failure.Furthermore, the existing GKE schemes require participants to perform all computations, which can degrade the performance of resource-constrained devices such as Internet of Things (IoT) devices.To solve these problems, in this paper, we propose an asynchronous GKE scheme that uses blockchain and smart contracts to store the security keys-related material and reduce the computational load of the participants.Furthermore, our proposed scheme provides Perfect Forward Secrecy (PFS) and Post-Compromised Security (PCS).Our implementation on Ethereum shows that the proposed scheme can scale to more than 100 participants when combined with a distributed storage system.

Digital Twins (DTs) have the potential to revolutionize the healthcare industry by offering effective and efficient healthcare services. However, there is a lack of clarity regarding the challenges pertaining to the creation of healthcare DTs from the components, design, and operational requirements perspective. In this paper, we address these gaps by examining the healthcare challenges and presenting the groundbreaking applications of DTs in healthcare to identify how they can address the healthcare challenges. We then discuss healthcare DTs’ design and operational requirements, focusing on considerations that ensure their effective implementation and adoption. In accordance with the requirements, we propose an architecture for implementing healthcare DTs, covering the steps of building DTs from data collection and integration to DT model creation and deployment. Moreover, we present the proof of concept as a prototypical implementation of healthcare DT. Finally, we discuss the open challenges of healthcare DTs and future research directions.

Federated Learning (FL) is a distributed Machine Learning paradigm that allows multiple clients to collaboratively train a model under the control of a central server while keeping data locally in edge devices. To simplify workload management in FL ecosystems, cloud computing and container-based approaches such as Kubernetes (K8s) have been proposed for scalable deployment. Nonetheless, K8s can violate fundamental FL privacy principles, e.g., the inherent flat networking approach in K8s can potentially allow FL clients to access other client or domain resources. The latter poses an open research problem and gap in the literature because serious privacy risks can arise from attackers gaining access to any client in the FL setup. To address this problem, this paper presents a networking approach via network isolation at the link layer level, and authentication and data packet encryption at the network layer level. The former allows to create secure resource sharing, and the latter is used to protect in-transit data. For this purpose, we use a K8s networking operator and a secure network protocol suite. The above combination facilitates on-demand link-layer connectivity, per-link data source authentication, and confidentiality between FL actors. We tested our approach on a network testbed composed of different geo-located nodes where FL clients are deployed. Our promising results showcase the feasibility of the solution for privacy preservation at the network level in K8s-based FL.

Digital Twins (DTs), being the virtual replicas of their physical counterparts, share valuable knowledge of the underlying physical processes and act as data acquisition and dissemination sources to Cyber-Physical System (CPS). Moreover, without obstructing the ongoing operations, DTs also provide an assessment platform for evaluating the operational behavior and security of the CPS. Therefore, they become a potential source of data breaches and a broad attack surface for attackers to launch covert attacks. To detect and mitigate security loopholes in DTs, one of the potential solutions is to leverage a gamification approach that can assess the security level of DTs while providing security analysts with a controlled and supportive virtual training environment. Artificial Intelligence/Machine Learning (AI/ML)-based approaches can complement the idea of security orchestration and automation in the gamification approach. However, AI/ML-based DTs security solutions are generally constrained by the lack of transparency of AI operations, which results in less confidence in the decisions made by the AI models. To address the explainable security challenges of DTs, this article proposes a gamification approach called sEcuriNg dIgital twins through GaMification Approach (ENIGMA). While leveraging DTs as an offensive security platform, ENIGMA provides gaming scenarios to assess DTs' security and train security analysts. The game players within ENIGMA are humans (the attacker team) and AI agents (the defender team). Furthermore, ENIGMA is supported by an eXplainable AI (XAI)-based DT security assessment model that explains the decisions made based on the SHAP values by the AI model on attack vectors for the defender team, i.e., the AI agent. The SHAP values illustrate the contribution of different features towards predicting the outcome of attack vectors. This explanation can help security analysts to take security measures based on reasoned and trustworthy decisions. Finally, experimental validation has been carried out to demonstrate the viability of ENIGMA.

Conventional manufacturing systems are shifting toward smart manufacturing where a wide range of devices are connected to the Internet, thanks to advances in Internet of Things (IoT) technology. The high connectivity of such devices introduces security risks because malicious nodes may attempt to compromise or tamper with data generated by the devices. This in turn highlights the security and data trust challenges in smart manufacturing. In this chapter, we discuss blockchain as a solution to address the outlined challenges. We study the state-of-the-art solutions proposed in the literature to enhance security and establish trust in manufacturing. Next, we introduce a blockchain-based framework to secure smart manufacturing by analyzing the data logs produced by the devices. Ultimately, we envision a blockchain-based framework to enforce trustworthiness in Digital Twins (DTs) that mirror every facet of the industrial processes to help achieve predictive maintenance.

Efficient energy management of Distributed Renewable Energy Resources (DRER) enables a more sustainable and efficient energy ecosystem. Therefore, we propose a holistic Energy Management System (EMS), utilising the computational and energy storage capabilities of nearby Electric Vehicles (EVs), providing a low-latency and efficient management platform for DRER. Through leveraging the inherent, immutable features of Distributed Ledger Technology (DLT) and smart contracts, we create a secure management environment, facilitating interactions between multiple EVs and energy resources. Using a privacy-preserving load forecasting method powered by Vehicular Fog Computing (VFC), we integrate the computational resources of the EVs. Using DLT and our forecasting framework, we accommodate efficient management algorithms in a secure and low-latency manner enabling greater utilisation of the energy storage resources. Finally, we assess our proposed EMS in terms of monetary and energy utility metrics, establishing the increased benefits of multiple interacting EVs and load forecasting. Through the proposed system, we have established the potential of our framework to create a more sustainable and efficient energy ecosystem whilst providing measurable benefits to participating agents.