Ongoing research and industrial exploitation of SDN and NFV technologies promise higher flexibility on network automation and infrastructure optimization. Choosing the location of Virtual Network Functions is a central problem in the automation and optimization of the software-defined, virtualization-based next generation of networks such as 5G and beyond. Network services provided for autonomous vehicles, factory automation, e-health and cloud robotics often require strict delay bounds and reliability constraints influenced by the location of its composing Virtual Network Functions. Robots, vehicles and other end-devices provide significant capabilities such as actuators, sensors and local computation which are essential for some services. Moreover, these devices are continuously on the move and might lose network connection or run out of battery, which further challenge service delivery in this dynamic environment. This work tackles the mobility, and battery restrictions; as well as the temporal aspects and conflicting traits of reliable, low latency service deployment over a volatile network, where mobile compute nodes act as an extension of the cloud and edge computing infrastructure. The problem is formulated as a cost-minimizing Virtual Network Function placement optimization and an efficient heuristic is proposed. The algorithms are extensively evaluated from various aspects by simulation on detailed real-world scenarios.

With the increasing adoption of intelligent transportation systems and the upcoming era of autonomous vehicles, vehicular services (such as remote driving, cooperative awareness, and hazard warning) will have to operate in an ever-changing and dynamic environment. Anticipating the dynamics of traffic flows on the roads is critical for these services and, therefore, it is of paramount importance to forecast how they will evolve over time. By predicting future events (such as traffic jams) and demands, vehicular services can take proactive actions to minimize Service Level Agreement (SLA) violations and reduce the risk of accidents. In this paper, we compare several techniques, including both traditional time-series and recent Machine Learning (ML)-based approaches, to forecast the traffic flow at different road segments in the city of Torino (Italy). Using the most accurate forecasting technique, we propose n-max algorithm as a forecast-based scaling algorithm for vertical scaling of edge resources, comparing its benefits against state-of-the-art solutions for three distinct Vehicle-to-Network (V2N) services. Results show that the proposed scaling algorithm outperforms the state-of-the-art, reducing Service Level Objective (SLO) violations for remote driving and hazard warning services.

Thanks to network slicing, mobile networks can now support multiple and diverse services, each requiring different key performance indicators (KPIs). In this new scenario, it is critical to allocate network and computing resources efficiently and in such a way that all KPIs targeted by a service are met. Accounting for all sorts of KPIs (e.g., availability and reliability, besides the more traditional throughput and latency) is an aspect that has been scarcely addressed so far and that requires tailored models and solution strategies. We address this issue by proposing a novel methodology and resource orchestration scheme, named OKpi, which provides high-quality decisions on VNF (Virtual Network Function) placement and data routing, including the selection of radio points of attachment. Importantly, OKpi has polynomial computational complexity and accounts for all KPIs required by each service, and for any resource available from the fog to the cloud. We prove several properties of OKpi and demonstrate that it performs very closely to the optimum under real-world scenarios. We also implement OKpi in a testbed supporting a robot-based, smart factory service, and we present some field tests that further confirm the ability of OKpi to make high-quality decisions.

This paper proposes a dynamic pricing and revenue-driven service federation strategy based on a Deep Q-Network (DQN) to instantly and automatically decide federation across different service provider domains, each introduces dynamic service prices offering to its customers and towards other domains. A dynamic pricing model is considered in this work based on the analysis of real pricing data collected from public cloud provider, and upon this a dynamic arrival process as a result of the price changes is proposed for formulating the service federation problem as a Markov Decision Problem (MDP). In this work, several reinforcement learning algorithms are developed to solve the problem, and the presented results show that the DQN method reached 90% of the optimal revenue and outperformed existing state-of-the-art strategies, and it can learn the federation pricing dynamics to make optimum federation decisions according to price changes.

Real-time control of connected industry devices such as mobile robots constituting Industrial Internet of Things (IIoT) has made time-sensitive communications over Wi-Fi increasingly important. In order to provide support for time-sensitive services over Wi-Fi, key features of the IEEE 802.11 standard, such as restricted target wake time (rTWT) and multiuser orthogonal frequency division multiple access (OFDMA), can be exploited. However, even with rTWT and OFDMA, time-sensitive networking (TSN) over Wi-Fi is still a challenge given the unpredictability of wireless channels and the increasing number of Wi-Fi-enabled devices. In this article, we present a comprehensive network calculus-based analysis of the delay bounds achievable in Wi-Fi networks, leveraging these IEEE 802.11 enhancements alongside synchronized priority queuing via IEEE 802.1Qbv. We propose PONTE, a novel fully polynomial time approximation scheme (FPTAS) scheduler that guarantees strict nonviolation probabilities (e.g., 99.99%) for inelastic, time-critical traffic over an 802.11 wireless network. Our extensive analysis and simulations in an IIoT scenario demonstrate that PONTE effectively manages dense, mixed traffic flows, achieving TSN objectives with minimal impact on fairness. Our approach is two orders of magnitude faster than the state-of-the-art.