Video streaming has emerged as a dominant mode of entertainment and information dissemination. However, traditional approaches to video bitrate adaptation often fall short in non-stationary network environments, leading to suboptimal user experiences. To address these challenges, this work presents an innovative method leveraging Proximal Policy Optimization (PPO), a Reinforcement Learning (RL) algorithm, to optimize the Quality of Experience (QoE) in video streaming scenarios. A Network Digital Twin (NDT) environment, which allows to emulate production environments, and the end, facilitate high-fidelity training and evaluation while completely isolating live network operations from experimental risks. The PPO algorithm learns to make bitrate decisions based on a reward function that incorporates key QoE metrics, including blockiness and block loss, achieving an impressive 73.02% enhancement in overall performance. Experimental results demonstrate a significant reduction in block loss (66.87%) and an improvement in blockiness (6.40%), while also increasing the streaming bitrate (17.18%) from training to production deployment.

Open Radio Access Network (O-RAN) fosters a new networking paradigm that supports flexible and open architectures, while enabling the functional disaggregation of traditional, monolithic, base stations. Specifically, the Lower-Layer Split (LLS) option 7.2x divides the PHY layer into Low-PHY at Open -Radio Unit (O-RU) and High-PHY at Open -Distributed Unit (O-DU), leveraging the so-called open fronthaul network, which imposes stringent latency requirements. In this work, we introduce and thoroughly examine the time window relationships that define the latency budget allocated to the open fronthaul network. Additionally, we propose a novel methodology to analyze the delay experienced by a realistic traffic pattern over a spine-leaf topology, accounting for various O-DU locations and site pools. To address the increased traffic load introduced by the fronthaul, we present a theoretical model to characterize the average delay across different priority levels when open fronthaul traffic coexists with flows from other Radio Access Technologies (RATs). This model is then validated and broadened through ns-3 simulations. Our findings show the applicability of the proposed methodology, which could be exploited by network operators in the planning phase, evincing its suitability to obtain benchmark delay values, offering a reliable tool for performance evaluation in open fronthaul networks.

During the last years we have witnessed a growing trend towards the deployment of disaggregated Radio Access Network (RAN) networks. In this context, the Open-RAN (ORAN) architecture is one of the most relevant approaches for 5G/6G networks, and it is believed to be a key element of forthcoming deployments. The performance of the so called Open - Fronthaul ($\mathrm{O}-\text{FH}$) segment is of the utmost importance to ensure an appropriate overall system performance. In this sense, the overall delay of the communication between nodes within the O-RAN fronthaul needs to be kept at very low values, to fulfill the requirements of the corresponding technologies. In this paper we propose exploiting queuing theory to model the O-FH performance, including a realistic traffic characterization. In particular, we present a theoretical framework based on Open Jackson Networks to model the one-way end-to-end delay, and we compare it with a that obtained by a realistic simulation setup. The results show that the theoretical model provides a reasonably accurate behavior under certain conditions, setting in all cases a tight upper bound for the end-to-end delay, which we identify as a key performance indicator.

Novel networking paradigms based on programmability and softwarization are deeply transforming the architecture and end-to-end management of next-generation mobile networks. This demonstration provides a step beyond in this transformation process by showcasing the full cloud-native deployment of a disaggregated and distributed 5 G mobile network from core to Radio Access Network (RAN) using open-source software. The distinguishing features of the proposed approach are that the orchestration process includes the use of a cloud-based Distributed Unit (DU) communicating through O-RAN Split 7.2 with a Radio Unit (RU) performing over-the-air transmission, the end-to-end configuration of slices with on-demand activation of core network function components, and the dynamic configuration of transport network flows to differently route the traffic of active slices.

The development of techniques able to forecast the mobile network traffic in a city can feed data driven applications, as Virtual Network Functions (VNF) orchestrators, optimizing the resource allocation and increasing the capacity of mobile networks. Despite the fact that several studies have addressed this problem, many did not consider neither the traffic relationship among city regions nor the information retrieved from public transport stations, which may provide useful information to better anticipate the network traffic. In this paper, we propose a new deep learning based architecture to forecast the network traffic using representation learning and recurrent neural networks. The framework, named Mobile Traffic Predictor Enhanced by Neighboring and Transportation Data (MTP-NT), has two major components: the first one is responsible of learning from the time series of the region to be predicted, with the second one learning from the time series of both neighboring regions and public transportation stations. Several experiments were conducted over a dataset from the city of Milan, as well as comparisons against widely adopted and state-of-the-art techniques. The results shown in this paper demonstrate that the usage of public transport information contributes to improve the forecasts in central areas of the city, as well as in regions with aperiodic demands, such as tourist regions.

To address the challenges posed by the deployment of microservices of future end-user applications in the cloud continuum, a performance prediction model working together with a network elasticity controller will be needed. With that aim, this work introduces Layered Queuing-Graph Neural Networks (LQ-GNN), a novel Machine Learning (ML) approach to develop a generalized performance prediction model for microservicebased applications. Unlike previous works focused on individual applications, our proposal aims for a versatile model applicable to any microservice-based application, integrating the Layered Queueing Network (LQN) modeling with Graph Neural Networks (GNN). LQ-GNN allows to efficiently estimate the response time of applications under different resource allocations and placements on the computing continuum. The obtained evaluation results indicate that the proposed model achieves a prediction error below 10% when considering different evaluation scenarios. Compared to existing methodologies, our approach balances prediction accuracy and computational efficiency, making it viable for real-time deployments. Consequently, ML-based performance prediction can significantly enhance the resource management and elasticity control of microservice-based architectures, leading to more resilient and efficient systems.

Services rely more and more on cloud platforms to offer their products to end users. This implies that being able to estimate the latency required to reach those cloud platforms is of growing importance. To shed light on this crucial aspect, we perform a three-month measurement campaign involving traceroute measurements every 30 minutes over 256 pairs of source-destination probes, where the vantage points are located in different Cloud Service Providers (CSPs) and the destination probes belong to one of the main Infrastructure Operators (IOs) of Spain. We provide interesting insights obtained from analyzing the data resulting from this campaign. Among them, we observe that, as expected, distance is the unavoidable factor impacting cloud latency. Yet, other results are less anticipated, such as the great stability of the network, or the lack of performance difference when comparing standard and premium network service tiers. We also analyze the potential of forecasting the cloud latency both for future samples but also for unobserved connections.

The PREDICT-6G framework, a novel 6G network architecture aiming to provide reliable, predictable, and time sensitive networking across diverse applications, integrates a multi-technology, multi-domain network infrastructure, creating a homogeneous and deterministic E2E connectivity across wired and wireless elements. This architecture addresses the evolving needs of specialized applications, setting a new standard for future networking technologies. A key component is the AI-driven Multi-stakeholder Inter-domain Control-Plane, which ensures the seamless delivery of time-sensitive services.

The massive growth of live streaming, especially gaming-focused content, has led to an overall increase in global bandwidth consumption. Certain services see their quality diminished at times of peak consumption, degrading the quality of the content. This trend generates new research related to optimizing image quality according to network and service conditions. In this work we present a gaming streaming use case optimization on a real multisite 5G environment. The paper outlines the virtualized workflow of the use case and provides a detailed description of the applications and resources deployed for the simulation. This simulation tests the optimization of the service based on the addition of Artificial Intelligence (AI) algorithms, assuring the delivery of content with good Quality of Experience (QoE) under different working conditions. The AI introduced is based on Deep Reinforcement Learning (DRL) algorithms that can adapt, in a flexible way, to the different conditions that the multimedia workflow could face. That is, adapt, through corrective actions, the streaming bitrate, in order to optimize the QoE of the content on a real-time multisite scenario. The results of this work demonstrate how we have been able to minimize content losses, as well as the fact of obtaining high audiovisual multimedia quality results with higher bitrates, compared to a service without an optimizer integrated in the system. In a multi-site environment, we have achieved an improvement of 20 percentage points in terms of blockiness efficiency and also 15 percentage points in block loss.

This paper provides an extensive analysis of Reliable and Available Wireless (RAW) enhancements and solutions needed to manage industrial environments more effectively. Starting from the description of a representative industrial use case, an analysis of gaps and promising new extensions is performed. Namely, the need to (i) support multi-domain operation, at both technology and administrative levels; (ii) integrate RAW with edge architectures; and, (iii) increase the mobility support in RAW networks. The identified gaps are indeed not yet tackled by the relevant standardization development organizations, mainly the Internet Engineering Task Force (IETF), and are thus object of our future work.

The fifth-generation mobile evolution introduces Next-Generation Radio Access Networks (NG-RAN), splitting the RAN protocol stack into the eight disaggregated options combined into three network units, i.e., Central, Distributed, and Radio. The disaggregated units reach full interoperability on Open RAN. Further advances allow the RAN software to be virtualized (vNG-RAN) on top of general-purpose hardware, enabling the management of disaggregated units and protocols as radio functions. The placement of these functions is challenging since the best decision must be based on multiple constraints, e.g., the RAN protocol stack split, routing paths in network topologies with restricted bandwidth and latency, asymmetric computational resources, etc. The literature does not deal with general placement problems with high functional split options and protocol stack analysis. This article proposes the first exact model for positioning radio functions for vNG-RAN planning, named PlaceRAN, as a Binary Integer Linear Programming (BILP) problem. The objective is to minimize the computing resources and maximize the aggregation of radio functions. The evaluation considered two realistic network topologies, and the results reveal that PlaceRAN achieves an optimized high-performance aggregation level. It is flexible for RAN deployment overcoming the network restrictions, and up to date with the most advanced vNG-RAN design and development.

The fifth-generation mobile evolution enables Next-Generation Radio Access Networks (NG-RAN) transformations. The RAN protocol stack is split into eight disaggregated options combined in three network units, i.e., Central, Distributed, and Radio. Further advances allow the RAN functions to be virtualized on top of general-purpose hardware using the virtualized RAN (vRAN). The combination of NG-RAN and vRAN results in vNG-RAN, enabling the management of the disaggregated units and protocols as a set of radio functions. However, the orchestration-based placement of these radio functions is challenging since the best decision can be determined by multiple constraints involving RAN disaggregation, crosshaul network requirements, availability of computational resources, etc. This article proposes OPlaceRAN, a vNG-RAN deployment orchestrator framed within the NFV reference architecture and aligned with the Open RAN initiative. OPlaceRAN supports the dynamic placement of radio functions focusing on vNG-RAN planning and is designed to be agnostic to the placement optimization solution. We developed a prototype based on cloud-native tools to deploy RAN using containerized virtualization and the OpenAirInterface emulator. The evaluation is analyzed considering two different approaches as a proof-of-concept. First, we applied two placement solutions in a controlled real computing infrastructure with a crosshaul network. Second, we investigated the orchestrator’s scalability with a real and larger-scale topology. Our results show that OPlaceRAN is an effective cloud-native solution for containerized network function placement and agnostic to the placement solution, handling scale-out well. OPlaceRAN is up-to-date with the most advanced vNG-RAN design and development approaches, contributing to the evolution of fifth-generation networks.

Service providers highly rely on network softwarization for addressing the demanding use cases of the 5G verticals. In order for the 5G vertical scenarios to be seamlessly executed, where the coverage should be provided in large areas, the required infrastructure capacity and availability can only be supported by cross-operator cooperation. In this context, we employ a novel federated core-edge 5G architecture, where 5G vertical services are represented as service function chains (SFCs) and can be offered on the infrastructure owned either by the local operator or a foreign one. Furthermore, we propose two online SFC placement methods that aim to efficiently place the SFCs across the architecture, taking into account the deployment cost. A simulation of the optimal solution, based on an integer linear programming (ILP) approach, is evaluated against a hop-based heuristic placement method that runs on our experimental 5G platform. Our experimental results demonstrate that the optimized placement produces the most cost-effective solution. Our heuristic algorithm introduces a near-optimal solution, but with a higher deployment cost. Compared with the ILP approach, the heuristic solution provides lower complexity and execution time. Finally, in order to further augment the placement techniques, we propose enhanced lifecycle management methods that use live migration and scaling actions to further decrease the execution cost and adapt to real-time incoming traffic, respectively.

Edge cloud computing is a crucial enabler for applications or network services with stringent communication requirements in the 5G networks and beyond era. Although edge compute resources contribute to a significant portion of end-to-end performance (e.g., delay), the involved performance trade-offs in relation to the utilized virtualization technologies and application functions are not yet thoroughly investigated. In this context, we study the performance dynamics of alternative edge cloud technologies, including different unikernel flavors, container builds, and implementations of exemplary Web services. Our experiments do not target identifying a single best option, rather than the involved performance trade-offs with respect to both service and infrastructure viewpoints as well as realistic edge cloud tasks, including on service operation, cloud resource or service elasticity, dynamic resource allocation and removal. We provide insights gained from a relevant extensive comparative evaluation based on a novel edge cloud experimentation infrastructure. Our experimental results inspired a conceptual edge cloud orchestration platform and its basic design guidelines, i.e., mixing and matching heterogeneous edge cloud virtual entities with particular operational or performance requirements.

Isolation is one of the more relevant attributes associated to the idea of network slicing, introduced by 5G services. Through isolation it is expected that slices from different customers could gracefully coexist without interfering each other, in the sense that whatever misbehavior or unforeseen demand from one slice customer could not affect the communication service received by any other slice customer supported atop the same physical transport infrastructure. This paper surveys and compare different technical approaches that can be taken for providing distinct isolation levels in the transport network, as a major component of end-to-end network slices. Furthermore, a number of isolation feasibility indicators are defined and proposed. These indicators are based on the approaches referred before, as a mean of guiding orchestration decisions at the time of provisioning or reconfiguring the transport slices in the network.

The consumption of video contents is currently dominating the traffic observed in ISP networks. The distribution of that content is usually performed leveraging on CDN caches storing and delivering multimedia. The advent of virtualization is bringing attention to the CDN as use case for virtualizing the cache function. In parallel, there is a trend on sharing network infrastructures as a way of reducing deployment costs by ISPs. Then, an interesting scenario emerges when considering the possibility of sharing virtualized cache functions among ISPs sharing a common physical infrastructure, mostly considering that usually those ISPs offer similar content catalogues to final end users. This paper investigates through simulations the potential efficiencies that can be achieved when sharing a virtual cache function if compared to the classical approach of independent virtual caches operated per ISP.

This paper explores and makes a case for allocating a Wide-area Infrastructure Manager (WIM) on-demand to support softwarized network slicing, as part of the full NFVI virtualised infrastructure foundation, to ensure that the connectivity attributes prescribed to network slices can be managed with flexibility and adaptability in a full end-to-end slice. We show how creating a WIM on-demand and dynamically allocating a new WIM for each network slice, rather than having one for the whole network, can be beneficial for various slicing scenarios, in a similar way that a Virtual Infrastructure Manager (VIM) on-demand has been utilized. The paper considers some of the components, abstractions, and mechanisms of WIM on-demand.

The traditional telecommunications business is evolving towards offering a richer set of services beyond basic connectivity, leveraging on network programmability and virtualization. A versatile execution environment is required, capable of running different workloads in different locations in the network. Cloud computing is the key paradigm that allows fostering this trending change. One interesting question to solve is to what extent those computing environments have to move towards the network edge. Some services can be enabled by environments with increased capillarity, while others can be implemented in environments with more relaxed constraints (e.g., in terms of latency). This paper explores this topic by differentiating service edge from physical network edge and proposing an architecture based on the ALTO server for assisting orchestration systems in discriminating the suitable environments for each service. We present a network-flow strategy for assigning services to infrastructure elements following those precepts, together with an initial scalability evaluation of the proposed assignment solution to show its feasibility.

5G is the next step in the evolution of operational networks, happening after the successful introduction of past generations like 3G or 4G. However, different with previous evolutions, in the case of 5G there is a confluence of novel technologies at different levels, ranging from the radio access to the virtualization technologies. Building, running and operating all of these new innovations is extremely challenging because of the novelty and the lack of previous experience on integrating them altogether. The base that sustains these new capabilities is the emergence of softwarized networks enabling automation, programmability and virtualization. The paper addresses the case of deploying 5G service slices by describing an architecture design for enabling orchestration in a multi-site environment, formed by 5G end-to-end facilities offered to vertical industries for experimentation, providing realistic situations for all the stakeholders on such multi-site environments (i.e., customers, providers, manufacturers, integrators, etc).

`Roam Like at Home' is a slogan used to express the new roaming regulation applied to the countries within the European Union. Since June 15th, 2017, mobile network subscribers of the network operators within the European Union can use mobile services in other European countries without having extra charges for it. Although almost all mobile subscribers within countries of the European Union benefits from the new roaming rule, many mobile service providers are yet to be adapted to this change. Therefore, those providers bear the burden of the cost for the inter-provider traffic. Moreover, the latency between subscribers and their operator specific content must be reduced in order to provide a roaming service `like at home'. An emerging virtualization technology such as a Network Function Virtualization (NFV) sheds light on these challenges. This paper describes the concept of the mobile roaming service and our on-going implementation work as a part of 5G-PPP 5GEx project which aims at localizing inter-provider mobile traffic within the virtualized network infrastructure.

It is becoming clear that softwarization and virtualization of networking entail fundamental changes in how networking industry can control or troubleshoot traffic flows or enforce network-wide policies. Also application developers bystander major changes, with new opportunities for fast development of innovative applications and services. Network functions (NFs) are key building blocks to realise such potential. Indeed, NFs are currently subject of great development by diverse vendors for a large range of applications beyond traditional well known networking functionalities. Such new capabilities may benefit from flexible and scalable design as opposite to current blackbox view of NFs. In this paper we propose one such novel NF with flexible and scalable design. Our NF exploits the capabilities of the network coding (NC) concept to improve different metrics of service performance, hence resulting in a NC service that can be offered either to external customers or used as internal service by the network provider. Our main contribution is three-fold. First, we justify our approach of internal logical visibility of network coding functions (NCFs), which allows proprietary while interoperable NCFs. Second, we propose the modelling of NCFs internal logic in terms of their software functional architecture. Finally, we show that our functional logic naturally maps to softwarized network architectures, which in turn facilitates NCFs virtualization. We also discuss the applicability of our architectures to representative conceptual scenarios and identify next steps.

Next 5G networks will force service and network providers to support a huge variety of final services with very different requirements in terms of bandwidth, latency, etc. In parallel, vertical industries, as customers, will use 5G networks as technical enablers to support their businesses, demanding appropriate mechanisms for a flexible access and control of network and computing resource slices in the locations where such vertical have business. Since the footprint and the availability of resources to support the variety of services could be limited on the primary provider side, open environments enabling the trading of resources in the form of slices are required to facilitate the sharing of infrastructure with the necessary isolation and scalability. Here, this is exemplified by the proposal of adaptations between two prevalent architectures being defined in the EU H2020 projects 5G-Crosshaul and 5GExchange for allowing the trading of crosshaul resources enabling 5G services for Verticals.

The objectives of Cross Stratum Optimization (CSO) and SLA Management in multi-domain infrastructure services lead to challenges for maintaining autonomy of resource control and coordination of the SLA lifecycle. The GEYSERS project includes an approach and architecture for SLA Management to address these two problems. The conceptual foundations and blueprint of this architecture are presented and assessed.