Periodic Reporting for period 2 - MonB5G (Distributed management of Network Slices in beyond 5G) Reporting period: 2021-04-01 to 2023-04-30 Summary of the context and overall objectives of the project MonB5G is a 42-month project involving twelve organizations from eight European countries. It focuses on the smart, flexible, and automated management of 5G and 6G network resources using artificial intelligence (AI). The project implements AI-based mechanisms for zero-touch management and orchestration of network slices, enabling self-configuration, self-monitoring, self-healing, and self-optimization capabilities without human intervention. MonB5G project presented hierarchical, fault-tolerant, AI-based, automated network management framework that includes security and energy efficiency techniques for orchestrating many parallel network slices. The concept of “network slicing” referred to the division of a virtual network infrastructure into multiple customized and independent segments and the suitable allocation, per case of use, of network resources to efficiently serve different requirements. Each network slice, assisted by AI, ensured the seamless operation of demanding applications that require high capacity and low latency. In addition to the provision of reliable and high-quality services, the use of AI enhanced infrastructure security and improved energy efficiency of networks. Overall, the MonB5G project's potential impacts can also extend beyond technical advancements and have the potential to generate socio-economic benefits, drive innovation, and positively influence the wider society by improving network efficiency, enhancing service quality, strengthening security, promoting sustainability, and contributing to economic growth for operators and service providers. The decentralized autonomic network management and orchestration proposed by MonB5G, will enable telecom and service providers to optimize their resource usage thus reducing both CAPEX and OPEX paves the way to offer new innovative 5G services at lower costs to consumers. Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far MoMajor achievements of the MonB5G project can be summarised as follows: The MonB5G project incorporated AI-driven and decentralized management and orchestration architecture, aiming to achieve ZSM for network slices. This involved the development of an AI-driven In-Slice Management (ISM) concept that reduced the number of external slice interfaces and separates slices' management plane. Additionally, a multi-domain orchestration framework provided a strong separation between different orchestration domains. In terms of security, privacy, and resiliency, the MonB5G project introduced an AI-based closed control loop framework to detect and mitigate attacks on network slices. It also explored the concept of Security orchestration and Security as a Service, emphasizing the importance of a trusted architecture for network slicing deployment. The project demonstrated the capabilities of the closed control loop in detecting and mitigating attacks through three specific use cases: in-slice mMTC DDoS attacks on AMF, aLTEr attacks involving traffic steering and VNF instantiation, and attack detection and mitigation on the Federated Learning (FL) training process. Regarding technology enablers, the MonB5G project defined novel end-to-end (E2E) slice Key Performance Indicators (KPIs) for monitoring the performance of network slices. It employs graph-based learning techniques for slice KPI prediction and explored the use of FL to reduce SLA violations in beyond 5G network slicing. AI-based intra and inter slice admission control mechanisms are also developed to optimize resource allocation and ensure efficient slice management. Additionally, MonB5G focuses on energy efficiency by developing decentralized cross-domain Energy Efficient DEs and implementing energy-saving techniques at the RAN and Edge. Furthermore, the MonB5G project contributed to the research community by publishing 5G datasets collected from its testbeds, providing valuable resources for future investigations and advancements in the field. MonB5G also contributed to relevant standard bodies (and groups therein) in ITU-T, 3GPP and ETSI and several dissemination and communication activities. The MonB5G consortium has published 29 journal and 37 conference papers in prestigious international journals and conferences throughout the duration of the project (5 more conference papers were accepted). Moreover, the contributions in three white papers by the 5G-PPP WGs and in the “Towards Natively Intelligent Networks” chapter of the 6G architecture book titled "Towards Sustainable and Trustworthy 6G: Challenges, Enablers and Architectural Design" illustrate its impact on the transition from 5G and beyond to a well-developed 6G architecture. The MonB5G consortium participated in a variety of outreach events, including presentations, workshops, tutorials, and demonstrations, which are detailed in Deliverable D7.7. MonB5G has also filed 5 patents during its project execution. Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far) MonB5G demonstrates the integration of connectivity, storage, and computing resources, allowing telecom/ISP providers to adopt new service models. By aligning with the ETSI ZSM vision, the project showcases the hosting of virtual reality applications in different domains, with cross-domain end-to-end SLA coverage and elastic end-to-end slice lifecycle management. To address network scalability challenges, MonB5G develops a scalable monitoring system that minimizes measurement overhead. It also employs data-driven methods for optimal slice configuration, distributed architecture with local entities instantiated in edge domains, and a local security orchestrator for threat identification and response.Secure and trusted environments for virtualized networks are crucial, and MonB5G leverages AI techniques and emerging technologies to create a trustworthy infrastructure. It employs federated learning, blockchain, SECaaS, SDN, NFV, and VNF security to enhance trust and isolation between different tenants while mitigating security threats. The project proposes a zero-touch slicing design with autonomic, cognitive, and closed-loop management and orchestration. By embedding analytics and decision entities in domain and inter-domain managers, MonB5G achieves automatic assignment, scaling, migration, and optimization of slice-level resources.A significant goal of MonB5G is to reduce network energy consumption by ten times. It achieves this through energy-aware AI-based techniques for resource allocation and VNF placement, allowing for the deactivation of non-utilized network elements. Telecom operators and vendors may implement MonB5G's innovations, enhancing their products, solutions, competitiveness, and market reach.In terms of the European economy, MonB5G aims to accelerate 5G deployment by providing faster and more secure provisioning of innovative services. Market opportunities such as mobile network slicing, AI/automation for network management, and cybersecurity solutions benefit telecom vendors, operators, and vertical industries, supporting their digital transformation initiatives. For consumers and society, MonB5G's decentralized autonomic network management and orchestration optimize resource usage and improves security and resilience, leading to lower costs. This enables the provision of new innovative 5G services across various sectors, generating positive societal and environmental impacts. MonB5G architecture
