6G ecosystems provide the promise of marketplaces where multiple stakeholders can interact to buy and sell softwarized resources. As the use of such resources necessitates that buyers have access to the sellers' networks, e.g. on-premises, edge, or cloud, sellers will prefer trusted buyers. To ensure its vibrancy, it is imperative that new entrants are incentivized to join the marketplace, while existing stakeholders are also treated fairly to retain their trust. As reputation can serve as a proxy for trust, we explore the feasibility of a decentralized blockchainbased reputation engine, with a focus on the computation costs of different reputation initialization approaches. To this end, we have implemented such an engine via smart contracts in the Ethereum blockchain, using the associated Remix simulator to calculate gas (computation) costs, and have evaluated the tradeoffs among our initialization approaches. We further discuss the implications of our results on 6G decentralized marketplace governance policies.

6G provides the promise of sophisticated services dynamically created through the composition of softwarized resources such as holograms, sensing and related vehicular technology, and advanced remote medicine, to name a few. Therefore, within the 6G ecosystem there will be multitudes of heterogeneous stakeholders, such as providers of specialized services, application developers, as well as communication providers, wanting to interact with each other, e.g., to negotiate the use of such resources. Furthermore, stakeholders will be able to join or leave the 6G ecosystem at will. Given the scale of 6G, it will be necessary for stakeholders to be able to buy/sell such softwarized resources in real-time without pre-existing static trust relationships among them, and a variety of trust relationships will need to be supported. At the same time, to encourage marketplace growth, it will be essential to prevent new entrants from being starved out due to the lack of prior trust history. In this paper, we use game theory as an underlying basis for incentivizing and enforcing socio-economic marketplace policies such as social welfare and social fairness. We then demonstrate using formal verification how game-theoretic strategies can be automatically generated to use as a basis for implementation of such policies, and how cooperation among stakeholders can lead to better outcomes w.r.t. non-cooperative approaches in a model of a marketplace with varying trust relationships.

Transfer learning has emerged to address various challenges in machine learning by improving the efficiency and effectiveness of the learning process. We explore transfer learning in the context of reinforcement learning and apply it to optimize the resource allocation for network slicing for different industrial environments. We propose two transfer learning techniques and investigate whether the knowledge gained by a pre-trained deep reinforcement learning agent for an industrial environment can be leveraged when the environment changes. Experiments demonstrate that our proposed approach yields higher performance when compared with the training of a new agent for the changed environment by reducing training convergence time, improving performance stability and achieving higher reward.

Industrial scenarios comprise multiple devices executing periodic, mutually-dependent tasks with challenging communications requirements. 5G technology and RAN slicing make it possible to accommodate such requirements. The predictability of the environment can be exploited to improve the network efficiency and performance. We leverage Deep Reinforcement Learning to design a "production-aware" agent based on the Deep Deterministic Policy Gradient algorithm. The proposed scheme combines production and network information to select the spectrum configuration of each slice. In details, by exploiting the knowledge about the upcoming production tasks, the agent can effectively predict the required per-slice spectrum consumption in order to boost the service provisioning reliability and limit the spectrum over-provisioning. We compare the performance of this approach with a "production-unaware" agent and with the optimal spectrum allocation. Simulations show how our solution provides a per-slice reliability in terms of meeting the latency requirements higher than the one provided by the "production-unaware" agent. Moreover, it ensures a tight approximation of the optimal slice spectrum allocation.

Network slicingmight radically change the relations among different actors of the telecommunications ecosystem, where new players, active in different markets, could benefit of tailored connectivity services based on different business strategies. We argue that for fully exploiting the opportunities offered by network slicing, dynamic sharing of resources is crucial not only for efficiency and cost savings, but also for enabling a resource negotiation that can unleash the potential of new business relations. We develop an automated mechanism that allows tenants to take strategic decisions to optimize the management of their slices based on their instantaneous demands and model their interaction as in marketplace. We integrate our solution, based on game theory, on a 3GPP calibrated system level simulator, where a slice-aware scheduler enforces the tenants’ decisions at the Nash Equilibrium (NE). We compare our proposal with a static baseline, that assigns a fixed share of resources to each slice, and show that, by dynamically trading resources in the market, tenants achieve lower costs, and, therefore, higher profits. We provide an algorithmic implementation that guarantees the convergence to a single NE and test the computational complexity of our algorithm to an increasing number of slices in the system.

Multi-connectivity is emerging as a promising solution to provide reliable communications and seamless connectivity for the millimeter-wave frequency range. Due to the blockage sensitivity at such high frequencies, connectivity with multiple cells can drastically increase the network performance in terms of throughput and reliability. However, an inefficient link scheduling, i.e., over and under-provisioning of connections, can lead either to high interference and energy consumption or to unsatisfied user's quality of service (QoS) requirements. In this work, we present a learning-based solution that is able to learn and then to predict the optimal link scheduling to satisfy users' QoS requirements while avoiding communication interruptions. Moreover, we compare the proposed approach with two base line methods and the genie-aided link scheduling that assumes perfect channel knowledge. We show that the learning-based solution approaches the optimum and outperforms the base line methods.

Accurate and efficient resource utilization predictions are of vital importance for the future generation of mobile wireless networks. By anticipating network resource demand, the operator can perform proactive resource allocation and predictive network control to improve network resource efficiency. In this paper, we exploit deep and transfer learning algorithms for multi-step resource utilization prediction in radio networks. In particular, we propose long short-term memory network-based architectures with transfer learning for the multi-step prediction task, in order to address scalability, computation time and data storage limitations of current implementations for large-scale networks. We carry out extensive experiments on a dataset collected from an LTE field network. When predicting physical resource block percentage utilization, our approach achieves state of the art results with root mean square error below 12 for a four-hour-ahead prediction, in half of the computation time required by deep learning methods without transfer learning.

5G introduces a broad range of new applications with diverse and demanding performance targets that require substantial investments on the radio access infrastructure. The network slicing model allows to share resources and to distribute responsibilities for network capacity expansion to infrastructure providers and for identifying resource needs to slice tenants specialized in different application domains. In this paper, we focus on the problem of taking capacity expansion decisions considering the traffic needs and performance requirements of network slices instantiated on the access infrastructure. We build on a techno-economic model that allows to describe the trade of resources between infrastructure provider and tenants. The model is based on a resource market that generates revenues for matching infrastructure recurrent costs, but also extra revenues for capacity expansion and technology upgrade. Being the pricing scheme based on resource scarcity in different geographical regions, we propose to guide capacity expansion decisions through an automatic algorithm based on utility functions and available budget generated by the market. Results show the effectiveness of the approach proposed.

By abstracting network resources and functions, network slicing allows to partition a single physical infrastructure into multiple logical independent networks, called slices, that can be tailored to the diverse needs of a wide range of application domains. In order to be independently managed by different players (tenants), slices need to be logically isolated from each other. As some of the applications envisioned for 5G networks are quite critical from performance point of view, a static slicing of resources, where isolation is not only logical but also physical, may appear a natural solution. However, the price to pay in terms of resource overprovisioning can be high, and therefore dynamic allocation strategies, based on optimized resource sharing, are alternative solutions. In this work, we propose an optimization framework to compare the two resource allocation schemes and define a metric to quantify the slicing gain. We show that the dynamic slicing outperforms the static approach and that the gain is significant even for stringent performance requirements, making dynamic slicing a viable solution also for critical applications.

Dynamically sharing network resources in a sliced multi-tenant network can provide cost efficient solutions that are able to guarantee specific service requirements for 5G networks and beyond. By automatizing the negotiations between tenants and infrastructure providers over the shared resources, it is possible to maximize the flexibility of the network in a very short time frame, thus increase efficiency. However, negotiating resources in a reactive manner can bring risks to the tenants due to traffic variations, and can also limit the gain in terms of spectral efficiency for the infrastructure provider. In this paper, we focus on how to exploit anticipatory strategies relying on predicted information on users' conditions in order to improve the efficiency of the proposed dynamic network slicing and trading framework. In particular, we analyze how to integrate a prediction algorithm into our scheme and analyze the techno-economic impacts of the anticipatory approach. Finally, we introduce a novel filtering algorithm to limit the impacts of prediction errors. Our results prove that using anticipatory strategies in dynamic negotiations and resource allocation increases tenants' utilities, while allows the infrastructure provider to accommodate more requests.

One of the disruptive innovations introduced by 5G networks is the opportunity for a new group of stakeholders to be actively involved in the management of network slices with the role of tenants. This allows to go beyond the user-centric QoS paradigm of 4G, and to include tools for handling the aggregate performance of multiple services and user groups and to focus on slice resource management, also at the new 5G NR interface. So far, research efforts have privileged a first solution based on the concept of isolation between slices. However, proposed solutions are not particularly efficient due to the loss of pooling gains, and not very reliable due to variable channel conditions that with slice limited resources make performance not easily predictable. We propose a slice management framework where the shared resources are negotiated by tenants in a real-time market based on slice instantaneous demands. Our model, based on game theory, allows tenants to optimize their service strategies acquiring resources when and where it is necessary, according to the level of quality and reliability requested by the specific traffic types they handle. In this paper, we focus on modeling the game theoretical framework and on characterizing its equilibria in a multi-tenant scenario.

Expanding the market of mobile network services and defining solutions that are cost efficient are the key challenges for next generation mobile networks. Network slicing is commonly considered to be the main instrument to exploit the flexibility of the new radio interface and core network functions. It targets splitting resources among services with different requirements and tailoring system parameters according to their needs. Regulation authorities also recognize network slicing as a way of opening the market to new players who can specialize in providing new mobile services acting as “tenants” of the slices. Resources can also be distributed between infrastructure providers and tenants so that they meet the requirements of the services offered. In this paper, we propose a model for dynamic trading of mobile network resources in a market that enables automatic optimization of technical parameters and of economic prices according to high level policies defined by the tenants. We introduce a mathematical formulation for the problems of resource allocation and price definition and show how the proposed approach can cope with quite diverse service scenarios presenting a large set of numerical results.

Relaying techniques for millimeter-wave wireless networks represent a powerful solution for improving the transmission performance. In this paper, we quantify the benefits in terms of delay and throughput for a random-access multi-user millimeter-wave wireless network, assisted by a full-duplex network cooperative relay. The relay is equipped with a queue for which we analyze the performance characteristics (e.g., arrival rate, service rate, average size, and stability condition). Moreover, we study two possible transmission schemes: fully directional and broadcast. In the former, the source nodes transmit a packet either to the relay or to the destination by using narrow beams, whereas, in the latter, the nodes transmit both the destination and the relay in the same timeslot by using a wider beam but with lower beamforming gain. In our analysis, we also consider the beam alignment phase that occurs every time, a transmitter node changes the destination node. We show the duration of how the beam is aligned, as well as position and a number of transmitting nodes, significantly affect the network performance. In addition, we discuss the impact of beam alignment errors and imperfect self-interference cancellation technique at the relay for full-duplex communications. Moreover, we illustrate the optimal transmission scheme (i.e., broadcast or fully directional) for several system parameters and show that a fully directional transmission is not always beneficial, but in some scenarios, broadcasting and relaying can improve the performance in terms of throughput and delay.

Network slicing is one of the main novelty of 5G systems and it is expected to radically change the approach of network operators to the support of vertical applications with specific and stringent performance requirements that can be managed by tenants. Static allocation of resources to slices is a straightforward approach to meet Service Level Agreements (SLAs), but it may lead to dramatic inefficiencies and high costs. On the other side, dynamic sharing can greatly improve resource utilization, but it requires a strict control of performance parameters so as to guarantee small congestion probability and smoothed degradation when it occurs. In this paper, we propose a utility-based approach that can properly map the slice customization strategies that tenants can use to address the different traffic requirements and their specific service management policies. We define resource scheduling mechanisms that allow differentiation among slices and service prioritization through slice-specific parameters. We model resource allocation as a mathematical programming problem and present numerical results that show the viability of the proposed approach.

By predicting changes in service requirements and the evolution of network states, anticipatory networking aims to improve the performance of decision making and optimization algorithms. In this paper, we propose a reliability-aware scheduling model which exploits information about the upcoming channel states of users to proactively optimize resource allocation. Along with the optimal formulation, proven to be NP-hard, a heuristic is also presented. The evaluation of the proposed optimization model and its corresponding heuristic is carried out considering both perfect and imperfect prediction of the users' channel states. Numerical results show that both models are robust against errors and that the decrease in reliability, when the length of the prediction window is equal to 25 ms, is less than 10%.

Multi-connectivity is emerging as promising solution to provide reliable communications and seamless connectivity at the millimeter-wave frequency range. Due to the obstacles that cause frequent interruptions at such high frequency range, connectivity to multiple cells can drastically increase the network performance in terms of throughput and reliability by coordination among the network elements. In this paper, we propose an algorithm for the link scheduling optimization that maximizes the network throughput for multi-connectivity in millimeter-wave cellular networks. The considered approach exploits a centralized architecture, fast link switching, proactive context preparation and data forwarding between millimeter-wave access points and the users. The proposed algorithm is able to numerically approach the global optimum and to quantify the potential gain of multi-connectivity in millimeter-wave cellular networks.

Maintaining service guarantees in a dynamic multi- tenant network, while ensuring an economically sustainable sharing platform, is a non-trivial problem. This paper, extending our previous work, develops a dynamic slicing and trading framework that can satisfy a variety of service guarantees. This framework not only determines the size of the network resource slices required for various active services, but it also adapts resource prices in accordance with the microeconomic laws of supply and demand. The proposed framework also ensures service continuity by learning the variations in the traffic mix as well as in the channel conditions, and by adjusting the slice assignments accordingly.

Flexible resource sharing at short time scales in multi-tenant shared radio access networks has proven to be quite a challenge. In this study, we develop a techno-economic model that enables dynamic short-term resource sharing as well as resource pricing, while simultaneously collecting revenue for network expansion. In order to regulate the resource costs and to prevent monopolization of resources, we define a unit cost of resources which can be scaled dynamically. The proposed framework allows operators to meet their individual utility targets while optimizing their expenditures based on their respective budgets. This work demonstrates that dynamic short timescale resource sharing can help network operators achieve their utility targets while minimizing their total expenditure.

Active sharing of wireless infrastructure can be an effective approach to reduce costs and improve network profitability. However, schemes proposed so far neither guarantee network operators the autonomy to compete and differentiate themselves in various market segments, nor give infrastructure providers the economic resources to keep the network updated in terms of technology and capacity. In this work, we propose a techno-economic model that allows network operators to compete and dynamically select the quality target to deliver to their customers, while simultaneously seeking to maximize their profits. In order to understand the willingness of network operators to participate in such a scenario, we develop a non-cooperative game wherein the Nash Equilibria show the propensity of operators to meet the customers' requirements. This work also points out the importance of retaining independent regulatory bodies, within the new business ecosystem, charged with proposing pricing policies capable of incentivizing investments towards infrastructure upgrades.

Communications using frequency bands in the millimeter-wave range can play a key role in future generations of mobile networks. By allowing large bandwidth allocations, high carrier frequencies will provide high data rates to support the ever-growing capacity demand. The prevailing challenge at high frequencies is the mitigation of large path loss and link blockage effects. Highly directional beams are expected to overcome this challenge. In this paper, we propose a stochastic model for characterizing beam coverage probability. The model takes into account both line-of-sight and first-order non-line-of-sight reflections. We model the scattering environment as a stochastic process and we derive an analytical expression of the coverage probability for any given beam. The results derived are validated numerically and compared with simulations to assess the accuracy of the model.

For the first time since the advent of mobile networks, the idea of advancing their pervasiveness by co-opting them into most aspects of daily life has taken hold and this idea is, henceforth, intended to be a mainstay of future networks (5G and beyond). As a result, a term one frequently encounters in the latest literature pertinent to radio access networks is reliability. It is, however, fairly evident that it is mostly used in a colloquial linguistic sense or that, in some cases, it is used synonymously with availability. This work is, to the best of our knowledge, the first to provide a quantitative definition of reliability which stems from its characterization in the dictionary and is based on quantifiable definitions of resilience, availability, and other parameters important to radio access networks. The utility of this quantitative definition is demonstrated by developing a reliability-aware scheduler which takes predictions of the channel quality into account. The scheduler developed here is also compared with the classical proportional fair scheduler in use today. This comparison not only succeeds in highlighting the practicality of the definition provided, but it also shows that the anticipatory reliability-aware scheduler is able to provide an improvement of about 35 - 50% in reliability when compared to a proportional fair scheduler which is common in contemporary use.

The exponential growth of media streaming traffic will have a strong impact on the bandwidth consumption of the future wireless infrastructure. One key challenge is to deliver services taking into account the stringent requirements of mobile video streaming, e.g., the users' expected Quality-of-Service. Admission control and resource allocation can strongly benefit from the use of anticipatory information such as the prediction of future user's demand and expected channel gain. In this paper, we use this information to formulate an optimal admission control scheme that maximizes the number of accepted users into the system with the constraint that not only the current but also the expected demand of all users must be satisfied. Together with the optimal set of accepted users, the optimal resource scheduling is derived. In order to have a solution that can be computed in a reasonable time, we propose a low complexity heuristic. Numerical results show the performance of the proposed scheme with respect to the state of the art.

Sharing mobile networks results in significant cost savings for operators, but limits competition due to loss of differentiation between operators. In this work we show that operators actually can differentiate even though they have to jointly take certain decisions, e.g., whether to activate or deactivate a shared small cell for energy efficiency purposes. The idea is to base the sharing agreement not on a fixed resource ratio, but to allow deviations from a more flexible baseline resource ratio that accommodate one operator's policy as long as the other operators' policies are not violated. We observe that activating additional small cells as a reaction to an increasing traffic demand can be delayed compared to the case with a strict resource ratio resulting in a significant energy saving.

Sharing wireless channel resources among cellular network operators saves costs by efficiently utilizing hardware and spectrum. Following the approach of resource sharing, we introduce a general formulation that covers current fixed and dynamic approaches as special cases. Unlike previous approaches, our formulation maximizes spectral efficiency by occasionally penalizing an operator within defined boundaries. This interesting tradeoff between sharing guarantees and system performance is studied by simulation. Our results clarify the costs of fair sharing and how much spectral efficiency can be gained by relaxing fairness in a controlled manner. This study, thus, provides important insight into the efficient design of shared base stations in future cellular networks.

Sharing wireless channel resources among cellular network operators saves costs by efficiently utilizing hardware and spectrum. Following the approach of Generalized Resource Sharing, this paper provides a mathematical rigorous formulation of multi-operator scheduling. The presented scheduling policy allows to trade off sharing guarantees versus spectral efficiency, covering current fixed and dynamic approaches as special cases. Analyzing this general scheduling policy leads to a profound understanding of its most important parameters and of their effect on rate-dependent utilities. The presented proofs are confirmed by simulation results that hold for insightful scenarios and point to promising future work.

Wireless access networks are often characterized by the interaction of different end users, communication technologies, and network operators. This paper analyzes the dynamics among these "actorsâ by focusing on the processes of wireless network selection, where end users may choose among multiple available access networks to get connectivity, and resource allocation, where network operators may set their radio resources to provide connectivity. The interaction among end users is modeled as a noncooperative congestion game, where players (end users) selfishly select the access network that minimizes their perceived selection cost. A method based on mathematical programming is proposed to find Nash equilibria and characterize their optimality under three cost functions, which are representative of different technological scenarios. System level simulations are then used to evaluate the actual throughput and fairness of the equilibrium points. The interaction among end users and network operators is then assessed through a two-stage multileader/multifollower game, where network operators (leaders) play in the first stage by properly setting the radio resources to maximize their users, and end users (followers) play in the second stage the aforementioned network selection game. The existence of exact and approximated subgame perfect Nash equilibria of the two-stage game is thoroughly assessed and numerical results are provided on the "qualityâ of such equilibria.

End users getting connectivity from wireless access networks want to choose the "best" access opportunity, that is, the best base station in cellular systems, and the best access point in WLANs. The metrics used to drive the association procedures are usually based on "local" parameters like the received signal strength, and the base station/access point's load. In case connectivity is provided through a Wireless Mesh Network the quality perceived by the user upon association depends also on "global" network-wide parameters. This paper studies the dynamics of network association in Wireless Mesh Networks by resorting to game theoretic tools. We show how the association problem can be formalized as a non-cooperative game in which end users selfishly play to minimize a perceived association cost which accounts for characteristics of the entire path to reach the WMN gateway. The quality of the Nash Equilibria for the proposed game are then quantitatively analyzed, and preliminary numerical results on the perceived association cost are derived for sample network topologies.

Cognitive radio networks aim at enhancing spectrum utilization by allowing cognitive devices to opportunistically access vast portions of the spectrum. To reach such ambitious goal, cognitive terminals must be geared with enhanced spectrum management capabilities including the detection of unused spectrum holes (spectrum sensing), the characterization of available bands (spectrum decision), the coordination with other cognitive devices in the access phase (spectrum sharing), and the capability to handover towards other spectrum holes when licensed users kick in or if a better spectrum opportunity becomes available (spectrum mobility). In this paper, a game theoretic framework is proposed to evaluate spectrum management functionalities in cognitive radio networks. The spectrum selection process is cast as a non-cooperative game among secondary users who can opportunistically select the "best" spectrum opportunity, under the tight constraint not to harm primary licensed users. Different quality measures for the spectrum opportunities are considered and evaluated in the game framework, including the spectrum bandwidth, and the spectrum opportunity holding time. The cost of spectrum mobility is also accounted in the analytical framework. Numerical results are reported to assess the quality of the game equilibria.

In future generation wireless access networks, the users will have the chance of choosing among multiple connectivity opportunities provided by different access networks (network selection problem). Moreover, the network operators themselves will have to implement effective resource allocation strategies taking wise decisions on the used technologies, frequencies, power levels, etc. This paper proposes a game-theoretic framework to model the problems of network selection and resource allocation, capturing the interdependencies of decisions taken by different players (users vs networks). Namely, we cast the problem as a non-cooperative game where users and access networks act selfishly according their specific objectives: maximization of the perceived quality of service for the end users, maximization of the number of customers for the access networks. We characterize the equilibria of the game by resorting to mathematical programming, and we derive numerical results to assess the "quality" of the equilibria.