Network slicing is a new paradigm for future 5G networks where the network infrastructure is divided into slices devoted to different services and customized to their needs. With this paradigm, it is essential to allocate to each slice the needed resources, which requires the ability to forecast their respective demands. To this end, we present DeepCog, a novel data analytics tool for the cognitive management of resources in 5G systems. DeepCog forecasts the capacity needed to accommodate future traffic demands within individual network slices while accounting for the operator's desired balance between resource overprovisioning (i.e., allocating resources exceeding the demand) and service request violations (i.e., allocating less resources than required). To achieve its objective, DeepCog hinges on a deep learning architecture that is explicitly designed for capacity forecasting. Comparative evaluations with real-world measurement data prove that DeepCog's tight integration of machine learning into resource orchestration allows for substantial (50% or above) reduction of operating expenses with respect to resource allocation solutions based on state-of-theart mobile traffic predictors. Moreover, we leverage DeepCog to carry out an extensive first analysis of the trade-off between capacity overdimensioning and unserviced demands in adaptive, sliced networks and in presence of real-world traffic.

As a chain is as strong as its weakest element, so as the efficiency, flexibility, and robustness of a mobile network, which relies on a range of different functional elements and mechanisms.. Indeed, the mobile network architecture needs particular attention when discussing the evolution of 3GPP EPS because it is the architecture which integrates the many different future technologies into one mobile network. This article discusses 3GPP EPS mobile network evolution as a whole, analyzing specific architecture properties which are critical in future 3GPP EPS releases. In particular, this article discusses the evolution towards a "network of functions," networking slicing, and software-defined mobile network control, management, and orchestration. Furthermore, the roadmap for the future evolution of 3GPP EPS and its technology components is detailed and relevant standards defining organizations are listed.

The emerging network slicing paradigm for 5G provides new business opportunities by enabling multi-tenancy support. At the same time, new technical challenges are introduced, as novel resource allocation algorithms are required to accommodate different business models. In particular, infrastructure providers need to implement radically new admission control policies to decide on network slices requests depending on their Service Level Agreements (SLA). When implementing such admission control policies, infrastructure providers may apply forecasting techniques in order to adjust the allocated slice resources so as to optimize the network utilization while meeting network slices' SLAs. This paper focuses on the design of three key network slicing building blocks responsible for (i) traffic analysis and prediction per network slice, (ii) admission control decisions for network slice requests, and (iii) adaptive correction of the forecasted load based on measured deviations. Our results show very substantial potential gains in terms of system utilization as well as a trade-off between conservative forecasting configurations versus more aggressive ones (higher gains, SLA risk).

Abstract-In addition to providing substantial performance enhancements, future 5G networks will also change the mobile network ecosystem. Building on the network slicing concept, 5G allows to "slice" the network infrastructure into separate logical networks that may be operated independently and targeted at specific services. This opens the market to new players: the infrastructure provider, which is the owner of the infrastructure, and the tenants, which may acquire a network slice from the infrastructure provider to deliver a specific service to their customers. In this new context, we need new algorithms for the allocation of network resources that consider these new players. In this paper, we address this issue by designing an algorithm for the admission and allocation of network slices requests that (i) maximises the infrastructure provider's revenue and (ii) ensures that the service guarantees provided to tenants are satisfied. Our key contributions include: (i) an analytical model for the admissibility region of a network slicing-capable 5G Network, (ii) the analysis of the system (modelled as a Semi-Markov Decision Process) and the optimisation of the infrastructure provider's revenue, and (iii) the design of an adaptive algorithm (based on Q-learning) that achieves close to optimal performance.

Software Defined Networking (SDN), characterized by a clear separation of the control and data planes, is being adopted as a novel paradigm for wired networking. With SDN, network operators can run their infrastructure more efficiently, supporting a faster deployment of new services while enabling key features such as virtualization. In this article, we adopt an SDN-like approach applied to wireless mobile networks that will not only benefit from the same features as in the wired case, but also will leverage on the distinct features of mobile deployments to push improvements even further. We illustrate with a number of representative use cases the benefits from the adoption of the proposed architecture, which is detailed in terms of modules, interfaces and high-level signaling. We also review the ongoing standardization efforts, and discuss the potential advantages, weaknesses and the need for a coordinated approach.

The virtualization of radio access networks (vRAN) is the last milestone in the NFV revolution. However, the complex relationship between computing and radio dynamics make vRAN resource control particularly daunting. We present vrAIn, a resource orchestrator for vRANs based on deep reinforcement learning. First, we use an autoencoder to project high-dimensional context data (traffic and channel quality patterns) into a latent representation. Then, we use a deep deterministic policy gradient (DDPG) algorithm based on an actor-critic neural network structure and a classifier to map contexts into resource control decisions. We have evaluated vrAIn experimentally, using an open-source LTE stack over different platforms, and via simulations over a production RAN. Our results show that: (i) vrAIn provides savings in computing capacity of up to 30% over CPU-agnostic methods; (ii) it improves the probability of meeting QoS targets by 25% over static policies; (iii) upon computing capacity under-provisioning, vrAIn improves throughput by 25% over state-of-the-art schemes; and (iv) it performs close to an optimal offline oracle. To our knowledge, this is the first work that thoroughly studies the computational behavior of vRANs and the first approach to a model-free solution that does not need to assume any particular platform or context.

This paper addresses the slicing of Radio Access Network (RAN) resources by multiple tenants, e.g., virtual wireless operators and service providers. We consider a criterion for dynamic resource allocation amongst tenants, based on a weighted proportionally fair objective, which achieves desirable fairness/protection across the network slices of the different tenants and their associated users. Several key properties are established, including: the Pareto-optimality of user association to base stations, the fair allocation of base stations' resources, and the gains resulting from dynamic resource sharing across slices, both in terms of utility gains and capacity savings. We then address algorithmic and practical challenges in realizing the proposed criterion. We show that the objective is NP-hard, making an exact solution impractical, and design a distributed semi-online algorithm which meets performance guarantees in equilibrium and can be shown to quickly converge to a region around the equilibrium point. Building on this algorithm, we devise a practical approach with limited computational, information, and handoff overheads. We use detailed simulations to show that our approach is indeed near-optimal and provides substantial gains both to tenants (in terms of capacity savings) and end-users (in terms of improved performance).

It is now commonly agreed that future 5G Networks will build upon the network slicing concept. The ability to provide virtual, logically independent "slices" of the network will also have an impact on the models that will sustain the business ecosystem. Network slicing will open the door to new players: the infrastructure provider, which is the owner of the infrastructure, and the tenants, which may acquire a network slice from the infrastructure provider to deliver a specific service to their customers. In this new context, how to correctly handle resource allocation among tenants and how to maximize the monetization of the infrastructure become fundamental problems that need to be solved. In this paper, we address this issue by designing a network slice admission control algorithm that (i) autonomously learns the best acceptance policy while (ii) it ensures that the service guarantees provided to tenants are always satisfied. The contributions of this paper include: (i) an analytical model for the admissibility region of a network slicing-capable 5G Network, (ii) the analysis of the system (modeled as a Semi-Markov Decision Process) and the optimization of the infrastructure providers revenue, and (iii) the design of a machine learning algorithm that can be deployed in practical settings and achieves close to optimal performance.