This article introduces the key innovations of the 5Growth service platform to empower verticals industries with an AI-driven automated 5G End-to-End (E2E) slicing solution which allows industries to achieve their service requirements. Specifically, we present multiple vertical pilots (Industry 4.0, Transportation and Energy), identify the key 5G requirements to enable them and analyze existing technical and functional gaps as compared to current solutions. Based on the identified gaps, we propose a set of innovations to address them with: (i) support of 3GPP-based RAN slices by introducing a RAN slicing model and providing automated RAN orchestration and control, (ii) an AI-driven closed-loop for automated service management with Service Level Agreement (SLA) assurance, and, (iii) Multi-domain solutions to expand service offerings by aggregating services and resources from different provider domains and also enable the integration of private 5G networks with public networks.

The increase of demand for mobile data services requires a massive network densification. A cost-effective solution to this problem is to reduce cell size by deploying a low-cost all-wireless Network of Small Cells (NoS). These hyperdense deployments create a wireless mesh backhaul amongst Small Cells (SCs) to transport control and data plane traffic. The semi-planned nature of SCs can often lead to dynamic wireless mesh backhaul topologies. This paper presents a self-organized backpressure routing scheme for dynamic SC deployments (BS) that combines queue backlog and geographic information to route traffic in dynamic NoS deployments. BS aims at relieving network congestion, whilst having a low routing stretch (i.e., the ratio of the hop count of the selected paths to that of the shortest path). Evaluation results show that, under uncongested conditions, BS shows similar performance to that of an Idealized Shortest PAth routing protocol (ISPA), while outperforming Greedy Perimeter Stateless Routing (GPSR), a state of the art geographic routing scheme. Under more severe traffic conditions, BS outperforms both GPSR and ISPA in terms of average latency by up to a 85% and 70%, respectively. We conducted ns-3 simulations in a wide range of sparse NoS deployments and workloads to support these performance claims.

Dense small cell (SC) deployments pose new architectural and transport level requirements. It is expected that they will require local multi-hop wireless backhauls in which some SCs with high capacity links towards the core act as aggregation gateways. To enable scalable deployments, there is the need of routing schemes able to handle dynamicity coming not only from increasing traffic demands and varying wireless link quality, but also from incremental gateway deployment. This letter proposes Anycast Backpressure (AB), a practical distributed anycast routing protocol designed to scale with the number of gateways and to exploit path and gateway diversity. The distinguishing feature of AB is that under increasing traffic demands, it opportunistically exploits uncongested gateways. To distribute traffic among the gateways, AB solely relies on 1-hop neighborhood queue backlog and geographic information. Ns-3 simulations show that, compared to state-of-the-art single-path and multi-path multi-gateway schemes, AB results in up to 40% gains in aggregated throughput and 99% reduction in latency for the evaluated scenarios.

The automated assurance of vertical service level agreements (SLA) is a challenge in 5G networks. The EU 5Growth project designs and develops a 5G End-to-End service platform that integrates Artificial Intelligence (AI) and Machine Learning (ML) techniques for any decision-making process in the management and orchestration (MANO) stack. This paper presents the detailed architecture and first prototype of the 5Growth platform taking AI/ML-based network service autoscaling decisions. This also includes the modification of the ETSI network service descriptors for requesting AI/ML-based decisions for orchestration problems and the integration of a data engineering pipeline for real-time data gathering and model execution.Our evaluation shows that AI/ML-related service handling operations (1-2 s.) are well below instantiation/termination procedures (80/60 s., respectively). Furthermore, online classification can be performed in the order of hundreds of milliseconds (600 ms).

Upcoming 5G mobile networks are addressing ambitious Key Performance Indicators (KPIs) not just in terms of capacity and latency, but also in terms of network control and management. In this direction, network management schemes need to evolve to provide the required flexibility, and automated and integrated management of 5G networks. This also applies to the 5G-Crosshaul transport network, which provides an integrated fronthaul and backhaul. Software Defined Networking (SDN) and Network Function Virtualization (NFV) are seen as key enablers for that. This article validates the flexibility, scalability, and recovery capabilities of the 5G-Crosshaul architecture in a testbed distributed geographically. More specifically, the central component of the validation is the hierarchical 5G-Crosshaul control infrastructure (XCI), conceived to handle multi-domain multi-technology transport network resources. Its performance is characterized through two experimental case studies. The first one illustrates the automated provisioning of all network resources required to deploy a complete LTE virtual mobile network featuring fronthaul and backhaul configurations. This takes 10.467s. on average for the network under test. The second one exploits the flexibility of the hierarchical XCI to apply local or centralized service recovery in the event of link failure depending on the desired path optimality vs. recovery time trade-off. On average, recovery takes 0.299s. and 6.652s., respectively. Overall, the proposed solution contributes to attain the target set for 5G networks of reducing service setup from hours to minutes.