Mobile networks are facing an unprecedented demand for high-speed connectivity originating from novel mobile applications and services and, in general, from the adoption curve of mobile devices. However, coping with the service requirements imposed by current and future applications and services is very difficult since mobile networks are becoming progressively more heterogeneous and more complex. In this context, a promising approach is the adoption of novel network automation solutions and, in particular, of zero-touch management techniques. In this work, we refer to zero-touch management as a fully autonomous network management solution with human oversight. This survey sits at the crossroad between zero-touch management and mobile and wireless network research, effectively bridging a gap in terms of literature review between the two domains. In this paper, we first provide a taxonomy of network management solutions. We then discuss the relevant state-of-the-art on autonomous mobile networks. The concept of zero-touch management and the associated standardization efforts are then introduced. The survey continues with a review of the most important technological enablers for zero-touch management. The network automation solutions from the RAN to the core network, including end-to-end aspects such as security, are then surveyed. Finally, we close this article with the current challenges and research directions.

Artificial Intelligence (AI) has become a major innovative force and a major pillar in the fourth industrial revolution. This trend has been acknowledged by the European Commission, who has pointed out how high-performance, intelligent, and secure networks are fundamental for the evolution of the multiservice Next Generation Internet (NGI). While great progress has been done in the accuracy and performance of AI-enabled platforms, their integration in autonomous decision-making and critical systems requires end-to-end quality assurance. AI@EDGE addresses these challenges harnessing the concept of “reusable, secure, and trustworthy AI for network automation”. To this end, AI@EDGE targets significant breakthroughs in two fields: (i) general-purpose frameworks for closed-loop network automation capable of supporting flexible and programmable pipelines for the creation, utilization, and adaptation of the secure, reusable, and trustworthy AI/ML models; and (ii) converged connect-compute platform for creating and managing resilient, elastic, and secure end-to-end slices supporting a diverse range of AI-enabled network applications. Cooperative perception for vehicular networks, secure, multi-stakeholder AI for Industrial Internet of Things, aerial infrastructure inspections, and in-flight entertainment are the uses cases targeted by AI@EDGE to maximise its commercial, societal, and environmental impact.

The complexity of wireless and mobile networks is growing at an unprecedented pace. This trend is proving current network control and management techniques based on analytical models and simulations to be impractical, especially if combined with the data deluge expected from future applications such as augmented reality. This is particularly true for software-defined wireless local area networks (SO-WLANs). It is our belief that to battle this growing complexity, future SO-WLANs must follow an artificial intelligence (AI) -native approach. In this article, we introduce aiOS, which is an AI-based platform that builds toward the autonomous management of SD-WLANs. Our proposal is aligned with the most recent trends in in-network AI promoted by the ITU Telecommunication Standardization Sector (ITU-T) and with the architecture for disaggregated radio access networks promoted by the Open Radio Access Network Alliance. We validate aiOS in a practical use case, namely frame size optimization in SD-WLANs, and we consider the long-term evolution, challenges, and scenarios for AI-assisted network automation in the wireless and mobile networking domain

Network Function Virtualization (NFV) and Multi-access Edge Computing (MEC) are two technologies expected to play a vital role in 5G and beyond networks. However, adequate mechanisms are required to meet the dynamically changing network service demands to utilize the network resources optimally and also to satisfy the demanding QoS requirements. Particularly in multi-domain scenarios, the additional challenge of isolation and data privacy among domains needs to be tackled. To this end, centralized and distributed Artificial Intelligence (AI)-driven resource orchestration techniques (e.g., virtual network function (VNF) autoscaling) are foreseen as the main enabler. In this work, we propose deep learning models, both centralized and federated approaches, that can perform horizontal and vertical autoscaling in multi-domain networks. The problem of autoscaling is modelled as a time series forecasting problem that predicts the future number of VNF instances based on the expected traffic demand. We evaluate the performance of various deep learning models trained over a commercial network operator dataset and investigate the pros and cons of federated learning over centralized learning approaches. Furthermore, we introduce the AI-driven Kubernetes orchestration prototype that we implemented by leveraging our MEC platform and assess the performance of the proposed deep learning models in a practical setup. 

While video-on-demand still takes up the lion's share of Internet traffic, we are witnessing a significant increase in the adoption of mobile applications defined by tight bit rate and latency requirements (e.g., augmented/virtual reality). Supporting such applications over a mobile network is very challenging due to the unsteady nature of the network and the long distance between the users and the application back-end, which usually sits in the cloud. To address these and other challenges, like security, reliability, and scalability, a new paradigm termed multi-access edge computing (MEC) has emerged. MEC places computational resources closer to the end users, thus reducing the overall end-to-end latency and the utilization of the network backhaul. However, to adapt to the volatile nature of a mobile network, MEC applications need real-time information about the status of the radio channel. The ETSI-defined radio network information service (RNIS) is in charge of providing MEC applications with up-to-date information about the radio network. In this article, we first discuss three use cases that can benefit from the RNIS (collision avoidance, media streaming, and Industrial Internet of Things). Then we analyze the requirements and challenges underpinning the design of a scalable RNIS platform, and report on a prototype implementation and its evaluation. Finally, we provide a roadmap of future research challenges.

The rollout of fifth generation (5G) cellular network technology has generated a new surge of interest in the potential of blockchain to automate various use cases involving cellular networks. 5G is indeed expected to offer new market opportunities for small and large enterprises alike. In this article, we introduce a new roaming network architecture for 5G based on a permissioned blockchain platform with smart contracts. The proposed solution improves the visibility for mobile network operators of their subscribers' activities in the visited network, as well as enabling quick payment reconciliation and reducing fraudulent transactions. The article further reports on the methodology and architecture of the proposed blockchain-based roaming solution using the Hyperledger platform. 

The ongoing roll-out of 5G networks paves the way for many fascinating applications such as virtual reality (VR), augmented reality (AR), and autonomous driving. Moreover, 5G enables billions of devices to transfer an unprecedented amount of data at the same time. This transformation calls for novel technologies like multi-access edge computing (MEC) to satisfy the stringent latency and bitrate requirements of the mentioned applications. The main challenge pertaining to MEC is that the edge MEC nodes are usually characterized by scarce computational resources compared to the core or cloud, arising the challenge of efficiently utilizing the edge resources while ensuring that the service requirements are satisfied. When considered with the users’ mobility, this poses another challenge, which lies in minimization of the service interruption for the users whose service requests are represented as service function chains (SFCs) composed of virtualized network functions (VNFs) instantiated on the MEC nodes or on the cloud. In this paper, we study the problem of joint user association, SFC placement, and resource allocation, employing mixed-integer linear programming (MILP) techniques. The objective function of this MILP-based problem formulation are to minimize (i) the service provisioning cost, (ii) the transport network utilization, and (iii) the service interruption. Moreover, a heuristic algorithm is proposed to tackle the scalability issue of the MILP-based algorithms. Finally, comprehensive experiments are performed to draw a comparison between these approaches.

In a world with increasing traffic demands, wireless technologies aim to meet them by means of new Radio Access Technologies that provide faster connectivity. Such is the case of 4G and 5G. However, in indoor scenarios, where the capabilities of these technologies are significantly affected by the distance to the base station and the materials used in the construction of buildings, Wi-Fi is still the technology of reference thanks to its low cost and easy deployment. In this context, it is usual to find multi-AP Wi-Fi networks whose deployment has been carefully planned. However, the user-AP association decision procedure is not defined by the IEEE 802.11 standard. As a result, vendors choose selfish approaches based on signal strength. This leads to uneven user distributions and nonoptimal resource utilization. To deal with this, densification has been used over the years, but this is expensive as it needs more infrastructure. Moreover, this results in more APs in the same collision domain. To avoid the need for densification, in this paper we introduce WiMCA, a joint SDN-based user association and channel assignment solution for Wi-Fi networks that considers signal strength, channel occupancy and AP load to make better association decisions. Experimental results have demonstrated that, in terms of aggregated goodput, WiMCA outperforms approaches based on signal strength by 55%, providing better user level fairness and accommodating more users and traffic before reaching the point at which densification is needed. © 2021, The Author(s)

Streaming high-quality video over dynamic radio networks is challenging. Dynamic adaptive streaming over HTTP (DASH) is a standard for delivering video in segments, and adapting its quality to adjust to a changing and limited network bandwidth. We present a machine learning-based predictive pre-fetching and caching approach for DASH video streaming, implemented at the multi-access edge computing server. We use ensemble methods for machine learning (ML) based segment request prediction and an integer linear programming (ILP) technique for pre-fetching decisions. Our approach reduces video segment access delay with a cache-hit ratio of 60% and alleviates transport network load by reducing the backhaul link utilization by 69%. We validate the ML model and the pre-fetching algorithm, and present the trade-offs involved in pre-fetching and caching for resource-constrained scenarios.

The exponential growth of mobile data traffic still remains an important challenge for the mobile network operators. In response, the 5G scene needs to couple fast connectivity and optimized spectrum usage with cloud networking and high processing power, optimally combined in a converged environment. In this paper, we investigate two 5G research projects; SESAME [1] and COHERENT [2]. We consider the proposed 5G architectures and the corresponding key network components, in order to highlight the common aspects towards the 5G architecture design.