Semantic and goal-oriented communication is increasingly studied for 6G, but generalization beyond seen data remains a key weakness under tight rate budgets. Many existing systems overfit their training data and degrade sharply at very low bit rates because they attempt to compress the entire signal. We introduce Foundation Model-Guided Semantic and Goal-Oriented Communication (FMSGOC), a framework that uses broad visual-linguistic Foundation Model priors to mitigate overfitting. It further improves rate efficiency by concentrating bits on sparse, goal-aligned anchors and relying on generative foundation-model priors to reconstruct the masked regions. By decoupling what to send from how to reconstruct, a vision-language foundation model selects and transmits a sparse set of semantic anchors, while a pretrained diffusion model, fine-tuned for masked completion, reconstructs the image at the receiver. In our experiments, FMSGOC reaches 0.039 bits per pixel (BPP), maintains high semantic fidelity (cosine similarity 0.87-0.90 on CIFAR-10), remains robust on previously unseen inputs (0.83-0.86 on ImageNet), and shows good perceptual similarity (0.1278/0.1558, CIFAR-10/ImageNet), outperforming strong end-to-end baselines at lower bit rates.

Goal-Oriented (GO) communication is an emerging paradigm that aims at enhancing the efficiency of Internet of Things (IoT) systems by sending only the minimum data needed to achieve the application goal. In GO communications, the network can apply techniques to reduce the amount of data sent through the network to a cloud node, such as analyzing and selecting data at the source with intelligent IoT devices. GO communication and the related networking techniques have primarily been studied from a device-centric perspective, often overlooking their impact on the network. However, the network consumes significant amounts of energy, with the Radio Access Network (RAN) alone accounting for approximately 70 % of the total energy consumption in mobile communication systems. Therefore, this paper introduces an end-to-end model for energy consumption, latency, and accuracy to assess GO networking strategies. We then use the model to evaluate the performance of GO strategies in edge cloud scenarios with different hardware platforms.

In future 6G networks, distributed management of network elements is expected to be a promising paradigm. Recent research progress in Artificial Intelligence (AI) is rapidly driving the adoption of distributed management. However, distributed management using intelligence or distributed AI inherently suffers from a number of issues - potential conflicts, signaling required to ensure cooperation and the convergence time of the algorithm. To this end, an early understanding and analysis of the overall effort to implement distributed AI in 6G, is still unexplored. This work, therefore, examines the impact of distributed AI, by analyzing its performance and how the existing 5G architecture could be enhanced to support it in 6G. We aim to understand the impact of distributed AI in 6G by selecting a relevant beyond 5G use case - auto-scaling virtual resources in a network slice. We present the performance and architecture analysis for two distributed algorithms from the domain of Reinforcement Learning - Q-Learning and Deep Q-Networks. We argue that despite its aforementioned issues, distributed AI brings benefits such as dynamic and adaptive decision-making, making it highly applicable for certain use cases in 6G.

The availability of networks able to perform control and management operations at unprecedented speeds is a crucial achievement for supporting next-generation Industry 4.0 applications. In particular, the presence of humans interacting with moving objects calls for decisions to be performed at time scales such as to ensure achieving the specific requirements identified by standardization bodies for the industrial sector in Beyond-5G environments. Among these specific Key Performance Indicators (KPIs), network reliability must be carefully pondered to guarantee the safety of the human operators at any time while investing the proper level of resources. For this reason, this paper presents an optimization problem that finds the best set of redundant radio bearers for each User Equipment (UE) in the factory area while accounting for the resource usage. Such problem extends the current specification on redundant transmissions. Several heuristics have been designed to achieve the desired time scales that could not be fulfilled with an exhaustive search. Results show that the optimization algorithm and the heuristics provide the same outcomes in terms of selected bearers and loss rates, but the latter do so at significantly more stringent time scales.

IT systems monitoring is a crucial process for managing and orchestrating network resources, allowing network providers to rapidly detect and react to most impediment causing network degradation. However, the high growth in size and complexity of current operational networks (2022) demands new solutions to process huge amounts of data (including alarms) reliably and swiftly. Further, as the network becomes progressively more virtualized, the hosting of NFV on cloud environments adds a magnitude of possible bottlenecks outside the control of the service owners. In this paper, we propose two deep learning anomaly detection solutions that leverage service exposure and apply it to automate the detection of service degradation and root cause discovery in a cloudified mobile network that is orchestrated by ETSI OSM. A testbed is built to validate these AI models. The testbed collects monitoring data from the OSM monitoring module, which is then exposed to the external AI anomaly detection modules, tuned to identify the anomalies and the network services causing them. The deep learning solutions are tested using various artificially induced bottlenecks. The AI solutions are shown to correctly detect anomalies and identify the network components involved in the bottlenecks, with certain limitations in a particular type of bottlenecks. A discussion of the right monitoring tools to identify concrete bottlenecks is provided.

Supporting end-to-end network slices and services across operators has become an important use case of study for 5G networks as can be seen by 5G use cases published in 3GPP, ETSI as well as NGMN. This paper presents the in- depth architecture, implementation and experiment on a multi-domain orchestration framework that is ab le to deploy such multi-operator service as well as monitor the service for SLA compliance. Our implemented architecture allows operators to abstract their sensitive details while exposing the relevant amount of information to support inter-operator slice creation. Our experiment shows that the implemented framework is capable of creating services across operators while fulfilling the respective service requirements.

While bringing many advantages, Software-Defined Networking (SDN) is accompanied by potential scalability issues that should be considered in the design of SDN-based networks. Specifically, SDN hardware switches based on Ternary Content-Addressable Memory (TCAM) can only store a few thousands of rules, imposing thus severe limits on the number of flows they can serve/process. Current proposals to deal with this problem mainly focus on optimal placement of flows that complies with the given constraints on TCAM size. We argue in this paper that flow routing not only should be but also can be independent of TCAM size constraints. We introduce two flow rule aggregation algorithms: One performs the "per-outport'' aggregation of the paths between access nodes in the network. It is optimal in that it holds the flow table sizes at the minimum, but has a drawback that it produces long identifiers of the path endpoints (access nodes), which cannot fit the IP address size. The other algorithm is its approximation under the constraint that the generated identifiers fit the limit imposed by the addressing scheme (e.g., IPv4). We study the performance of our algorithms analytically and through a set of experiments. While the optimal solution always keeps the flow table sizes at the minimum, we show that the approximate algorithm reduces the flow table sizes by a factor of 2 to 10 compared to the state of the art solution, under a reasonable constraint on the address length (e.g., 32 bits in case of IPv4).

For smaller scale wireless sensor networks (WSN) it has been clearly shown that a single mobile sink can be very beneficial with respect to the network lifetime. Yet, how to plan the trajectories of many mobile sinks in very large WSNs in order to simultaneously achieve lifetime and delay goals has not been treated so far. In this paper, we delve into this difficult problem and propose a heuristic framework using multiple orbits for the sinks' trajectories. The framework is carefully designed based on geometric arguments to achieve both, high lifetime and low delay. In simulations, we compare two different instances of our framework, one conceived based on a load balancing argument and one based on a distance minimization argument, with a set of different competitors spanning from statically placed sinks to battery-state aware strategies. We find our heuristics to perform very favorably: both instances outperform the competitors in both, lifetime and delay. Furthermore, and probably even more importantly, the heuristic, while keeping its good delay and lifetime performance, scales well with an increasing number of sinks.

The deficient energy supplies of wireless sensor networks (WSNs) drives network designers to optimize energy consumption in various ways. Not only with regard to the energy issue but also with respect to system performance, we design a local search technique for sink placement in WSNs that tries to minimize the maximum worst-case delay and extend the lifetime of a WSN, simultaneously. Since it is not feasible for a sink to use global information, which especially applies to large-scale WSNs, we introduce a self-organized sink placement (SOSP) strategy that combines the advantages of our previous works. The goal of this research is to provide a better sink placement strategy with a lower communication overhead. Avoiding the costly design of using nodes' location information, each sink sets up its own group by communicating to its n-hop distance neighbors. While keeping the locally optimal placement, SOSP exhibits a quality of the solutions with respect to communication overhead as well as computational effort that are better than previous solutions. To model and consequently control the worst-case delay of a given WSN we build upon the so-called sensor network calculus (a recent methodology first introduced).