The cloud-to-edge continuum paradigm has permeated various application domains, including critical urban-city safety systems. In these contexts, anomalies can compromise public safety, for example, by disrupting the communication between smart city infrastructure and vehicles, which aims to prevent accidents at pedestrian crossings. Given these environments’ heterogeneous and large-scale nature, manual recovery from anomalies is not feasible. Machine Learning techniques have emerged as an alternative, supporting a zero-touch approach that enables self-organising and self-healing solutions for anomaly prediction, detection, and mitigation. This paper proposes an Artificial Intelligence-driven, self-organising approach for anomaly management in the cloud-to-edge continuum, integrating both reactive and proactive mechanisms. We evaluate different Machine Learning models, including Random Forest Classifiers, Neural Networks, and Convolutional Neural Networks, to predict node performance anomalies. The simulation results obtained using the COSCO framework showcase the effectiveness of our method. It achieves an F1 score of 73

This demonstration presents a live, interactive showcase of RALA-PSO (Resource-and-Latency-Aware Particle Swarm Optimisation) algorithm [1] [2] integrated within the MIRO orchestration platform [3] for intelligent cloud-native scheduling. Attendees will experience hands-on interaction with three algorithmic variants that consider different fitness objectives via MIRO’s dashboard, observing live scheduling decisions and performance metrics integrated within a multidomain infrastructure. The demonstration illustrates state-of-the-art swarm intelligence research applied to a real Cloud-Native solution, bridging the gap between algorithmic innovation and practical deployment.

Swarm Intelligence (SI) has emerged as a prominent approach in resource scheduling, offering adaptability and effectiveness in multi-objective optimisation. However, SI-based methods often introduce complexity and challenges such as premature convergence and excessive exploration, leading suboptimal performance. To address these challenges, this paper presents a novel Resource-and-Latency-Aware Particle Swarm Optimisation (RALA-PSO) algorithm for intelligent and contextual cloud native resource scheduling. RALA-PSO expands upon Particle Swarm Optimisation (PSO) by incorporating a greedy initialisation for faster convergence, a circular search space to avoid boundary stagnation, a multi-criteria fitness function tailored to Cloud-Native characteristics, a heuristic-based scoring model to reward suitable solutions, and the integration of adaptive weights and learning factors for exploration-exploitation balance. The RALA-PSO implementation is available as open-source in [14]. RALA-PSO prioritise solutions based on efficient resource allocation and less communication overhead via service co-location. Experimental evaluations using the Alibaba Cluster TraceV2018 dataset show that RALA-PSO reduces network transmission overhead cost by 1.8x to 2.36x compared to state-of-the-art algorithms MOPPSO-CMS and ACO-CMS, while maintaining competitive resource efficiency. Notably, RALA-PSO achieves runtimes up to 212 times faster than ACO-CMS and 1.5 times faster than MOPPSO-CMS. These results demonstrate that RALA-PSO is a scalable, efficient, and context-aware scheduling solution for cloud-native applications, particularly effective in environments with strict network and resource constraints.

Service Function Chains allow the dynamic generation of basic virtual network services in order to build more complex services. They constitute the basis of the flexibility support for 5G and Beyond 5G networks. In order to totally exploit their benefits, the Service Function Chains should adapt to specific user requirements and current network status. The existing mechanisms to compose and embed Service Function Chains mostly disregard energy metrics to guide their decisions, thus resulting in a lack of energy optimisation by network operators. This paper presents a revision of existing Service Function Chain composition and embedding approaches and a discussion of current and future research directions, focused on energy optimisation for adaptive Service Function Chains. Furthermore, an architecture to empower the green orchestration of adaptive Service Function Chains is also drafted.

In the ever-evolving landscape of computing, the advent of edge and fog computing has revolutionized data processing by bringing it closer to end-users. While cloud computing offers numerous advantages, including mobility, flexibility and scalability, it introduces challenges such as latency. Fog and edge computing emerge as complementary solutions, bridging the gap and enhancing services’ proximity to users. The pivotal challenge addressed in this paper revolves around optimizing the placement of application microservices to minimize latency in the cloud-to-edge continuum, where a proper node selection may influence the app’s performance. Therefore, this task gains complexity due to the paradigm shift from monolithic to microservices-based architectures. Two distinct placement approaches, app-based and service-based, are compared through four different placement algorithms based on criteria such as link latency, node resources, and gateway proximity. App-based allocates all the services of one app sequentially, while service-based allocates one service of each app at a time. The study, conducted using YAFS (Yet Another Fog Simulator), evaluates the impact of these approaches on latency and load balance. The findings consistently confirm the hypothesis that strategies utilizing a service-based approach outperformed or performed equally well compared to app-based approaches, offering valuable insights into trade-offs and performance differences among the algorithms and each approach in the context of efficient microservices placement in cloud-to-edge environments.

The proliferation of automated services and applications has generated large amounts of data that must be handled with artificial intelligence and machine learning techniques, impacting businesses in different fields. Cognitive customer profiling services allow a better user experience enabling the tailoring of the customer experience to specific consumer patterns and preferences. Cognitive customer profiling services require heavy processing power in order to get accurate results from the artificial intelligence/machine learning models, which has led to services taking advantage of the cloud-to-edge continuum. This trending approach brings a set of challenges that have to be addressed. The more recent edge intelligence paradigm introduces artificial intelligence capacities to edge nodes, enabling a new set of solutions. However, artificial intelligence techniques must be adapted to deal with the specific characteristics of these scenarios. This paper identifies the challenges of running cognitive customer profiling in centralized scenarios, such as the cloud, to later introduce other approaches more suited for scenarios like the edge, identifying challenges and future research directions in the area.

Kubernetes is now the most widely used container orchestration tool in the Cloud. However, when deploying Kubernetes clusters in resource-constrained environments such as the Cloud-to-Edge continuum, new challenges arise. To address this issue, lightweight distributions of Kubernetes have been developed. It is crucial to fully understand the performance and security levels of the Kubernetes solution to deploy, as these factors could impact the services and applications running in the cluster. This research compares the performance and security of three Kubernetes distributions: K8s, K3s and K0s. Results indicated K3s lacks in performance due to scalability issues compared to K0s (top performer) and K8s. Moreover, the latter two exhibit fewer security vulnerabilities.

5G aims to support ubiquitous connectivity, ultra-Reliable Low Latency (uRLLC), and massive device communication in Next Generation networks. To achieve these objectives, the Open-Radio Access Networks (O-RAN) alliance aims to decouple the Radio Access Network (RAN) architecture and allow heterogeneity. To ensure the services' requirements, it is necessary to guarantee solutions that improve the management of the network. This work proposes an End-to-End (E2E) orchestration framework for a 5G communication infrastructure with open-source components. An overview of the implemented architecture is presented and two demonstrations are shown: how RAN and Core Network metrics are retrieved using a monitoring xAPP, and how the orchestrator enforces a policy after processing and analysing the data gathered. The results show that it is possible to deploy the proposed architecture to monitor and allocate resources efficiently in near-Real Time (near-RT) environments. The major novelty of this work is the fact that this constitutes the first E2E 5G network system using open-source tools, to the best of our knowledge. For this purpose, an interface adapter was built to interlink some of these open-source components.

As resource-demanding mobile applications become increasingly popular, smartphones, which are resource- and battery-dependent by nature, can migrate their workload to other devices on the Cloud-to-Edge continuum. This is known as computation offloading and allows for the “heavy” processing to be carried out on another, typically more powerful equipment that prompts the result back to the smartphone. The goal is to improve overall performance, reduce energy consumption, and/or prolong the smartphone's battery life. However, the offloading process might have practical implications, such as performance degradation by increasing the latency of the response time or even increasing the energy consumption of the device if the application requires heavy data transfer. Also, one should consider that the chips in today's smartphones are extremely energy efficient and offer outstanding performance. 5G networks also increase the data transfer bandwidth between devices. We aim to shed light on the circumstances under which computation offloading is a robust architectural solution for mobile apps. We used the EdgeBench benchmark in our experimental evaluation, namely the audio, image, and scalar applications, over three smartphones. The results highlight different performance and energy consumption depending on the type of device and manufacturer, considering the same application, showing that the offloading decision is not linear, thus it is not always the best solution to minimize battery consumption.

5G networks have strict constraints regarding the services in terms of latency, reliability, and availability, which pose additional challenges to the traditional orchestration solutions, and 6G networks will increment the number of slices and services deployed over different technological domains, adding more difficulties for the orchestrator. Distributed and automated solutions will be essential for this context. This article identifies the main challenges in 5G/6G orchestration and then describes the utility of Artificial Intelligence-driven solutions, outlining an orchestrator architecture for 5G networks. The architecture is then explored as an orchestration solution for two ongoing research projects focused on the deployment of critical services over 5G networks.

The Cloud-Fog-Internet of Things continuum combines different paradigms to provide connectivity and ubiquity for end-users, while also granting low latency and low jitter to cope with different challenges, including the requirements of latency-sensitive applications, such as virtual/augmented reality and online gaming. This constitutes a complex and dynamic environment with heterogeneous resources that need to be managed or orchestrated, in order to accomplish application requirements for low latency. Common orchestration solutions make placement decisions based only on the resources of the underlying network and the application resource requests; however, using the profiles of applications to make placement decisions has the potential to enhance the final performance perceived by the end-users. This paper proposes the use of application profiles according to their popularity to guide their placement. To corroborate the effectiveness of the use of the profiles, two placement mechanisms are presented, one based on Genetic Algorithm and the other inspired on graph partitions. Simulation results show that it is possible to reduce the latency and jitter of applications via a service placement guided by the profiles. The mechanism based on graph partitions showed better results for all scenarios, followed closely by the Genetic Algorithm in the scenarios with lower load.

The Internet of Things paradigm enables a new set of smart end-user applications. The Cloud-Fog-Mist-Internet of Things infrastructure provides communication, compute, and storage support for these applications. However, this complex, heterogeneous, and distributed landscape requires orchestration and management mechanisms in order to guarantee their proper functioning. One particular factor to manage is the capacity to provide service resilience even in the presence of failures in components of the substrate infrastructure. This research proposes a set of mechanisms to formalize, orchestrate, and embed a batch of service requests for chained Virtual Functions to fulfill the specific requirements of applications while enhancing their availability and ultimately their resilience. In detail, this work introduces a formal grammar to describe customized Service Chains, allowing the definition of replicas for different Virtual Functions, and an Integer Linear Programming model for Virtual Function embedding that prioritizes the use of nodes with higher availability. Additionally, an alternative heuristic is presented to handle more complex scenarios by taking advantage of the multi-tier scenario comprising the Cloud-Fog-Mist-Internet of Things. Simulation results for the embedding mechanisms show that it is possible to increase the resilience of chained Virtual Functions, while balancing the load of the infrastructure nodes.

As communications evolve to give space to new applications, such as augmented reality and virtual reality, new paradigms arise to provide essential characteristics like lower latency levels, mobility support, and location awareness. Such is the case of Fog computing, which extends from the well-known Cloud computing paradigm by bringing processing, communications, and storage capabilities to the edge of the network. By offering these novel features, also new challenges emerge that call for the design and implementation of orchestration mechanisms to deal with resource management. One of these mechanisms is related to the service placement, which consists in the selection of the appropriate execution node for the applications according to a specific optimization objective. In this paper, an Integer Linear Programming model for service placement aimed at latency reduction of popular applications is proposed. Furthermore, a heuristic based on the PageRank algorithm, called Popularity Ranked Placement, is also introduced. Simulation results show that the heuristic has lower execution times and is able to better balance the load in the network nodes, while being close to the ILP-based solution latency levels.

The Cloud to Fog continuum is a very dense and complex scenario. At the core level (Cloud) resources are vast, whilst they become scarce at the Edge (Fog). This complexity leads to the need of simulation tools in order to evaluate the performance of novel mechanisms that hardly can be tested in real scenarios. Thus, simulation represents a solution for early stage evaluation before moving to real-world (and more expensive and complex) testbeds. However, selecting the appropriate simulation tool can be complex in itself. This paper presents a conceptual review on six Cloud/Fog Simulation tools, describing their main characteristics and what they allow to experiment. A practical overview of the most representative Cloud/Fog simulators is presented, reporting about their resource consumption and execution time. The aim of this survey is to enlighten other researchers in the selection of the appropriate Cloud/Fog simulation tool for their goals and to know what they can expect from said tools.

With the advent of the Internet of Things many applications emerged that are not suitable for well-known paradigms like the Cloud, requiring its extension to provide more features to final users. Thus, the Fog rises as an extension to the Cloud able to provide mobility support, geographical distribution, and lower latency, by moving the services closer to the users, to the edge of the network. This new environment located at the edge comes with its own orchestration challenges. Among the orchestration functions that must be adapted to this new environment is scheduling. This paper presents a simple scheduling algorithm for Fog federative environments that organizes Fog instances into divisions for task assignment. Experimental results show that this approach could be particularly beneficial for critical time applications, commonly located at the Fog.

Fog computing provides a low latency access to resources at the edge of the network for resource-constrained devices. The high mobility of some of these devices, such as vehicles, brings great challenges related to resource allocation and management. In order to improve the management of computing resources utilized by mobile users connected to the Fog infrastructure, this paper proposes a virtual machine placement and migration decision model based on mobility prediction. Simulations have shown that moving the virtual machine to a Fog node ahead of the user's route using the proposed approach can decrease by almost 50 % the number of migrations needed by the user. The Fog architecture provides an average latency of about 15 milliseconds for the users' applications and the proposed approach presents a lower latency compared to a greedy approach for the VM placement problem.

Recent developments in telecommunications have allowed drawing new paradigms, including the Internet of Everything, to provide services by the interconnection of different physical devices enabling the exchange of data to enrich and automate people’s daily activities; and Fog computing, which is an extension of the well-known Cloud computing, bringing tasks to the edge of the network exploiting characteristics such as lower latency, mobility support, and location awareness. Combining these paradigms opens a new set of possibilities for innovative services and applications; however, it also brings a new complex scenario that must be efficiently managed to properly fulfill the needs of the users. In this scenario, the Fog Orchestrator component is the key to coordinate the services in the middle of Cloud computing and Internet of Everything. In this paper, key challenges in the development of the Fog Orchestrator to support the Internet of Everything are identified, including how they affect the tasks that a Fog service Orchestrator should perform. Furthermore, different service Orchestrator architectures for the Fog are explored and analyzed in order to identify how the previously listed challenges are being tackled. Finally, a discussion about the open challenges, technological directions, and future of the research on this subject is presented.

A new era of automated services has permeated user's daily lives thanks to paradigms such as Smart City and the Internet of Things. This shift from traditional applications is possible due to the massive amount of heterogeneous devices that constitute the Internet of Things. To provide newly improved characteristics to these services, such as mobility support, high resilience, and low latency, an extension to the Cloud computing paradigm was created, called Fog computing, which brings processing and storage resources towards the edge of the network, in the vicinity of the Internet of Things environment. This scenario implies a higher complexity level needed to coordinate available resources and how applications and services use them. Although some solutions have been proposed for the Cloud, several characteristics differentiate the Cloud from the Fog, creating the need for new mechanisms for the coordination of resources, applications, and services in the Fog. This paper explains the challenges present in the Fog that call for new mechanisms to later propose an architecture to manage resources in the Fog using a hybrid approach. In the Internet of Things and South-Bound Fog Levels, a distributed management of applications and services is proposed applying choreography techniques to enable automated fast decision making. A centralized approach to orchestrate applications and services taking advantage of a global knowledge of the resources available in the network is suggested for the North-Bound Fog and Cloud Levels.

The Internet of Things comprises a network of physical objects, like sensors and actuators, collecting and exchanging data. Given the importance of the information exchanged in these environments, the communication infrastructure becomes a critical point that needs to be managed optimally, while providing high-performance levels to end users. To guarantee the correct interaction between the different procedures intended for the optimization of the communication infrastructure, a standard and flexible representation of the data related to the network is necessary. Many of the services could be monitored through the web, thus using standard Web languages with a rich expressive power, such as the languages used in the Semantic Web, would allow for the reification of interoperable descriptions.This paper presents an ontology for the Internet of Things infrastructure tailored to the needs of Smart Cities. Furthermore, different kinds of evaluations were performed to corroborate the correctness of this ontology, including potential infrastructure optimization objectives like low latency and high resilience.

New services and applications become part of our daily activities as we evolve into new solutions supported by cutting-edge paradigms, like the Internet of Things and Smart Cities. In order to properly achieve the benefits theoretically provided by these models, new kinds of services must be designed. These new services have special requirements, as well as the users that access to them. One of these requirements is low latency levels, since a delayed reply could render to chaos for applications such as eHealth and public safety. The communication infrastructure must cope with these challenges by offering innovative solutions. One of these solutions is a smart service placement system that facilitates the location of services in the proper position according to specific needs. On this paper, a service placement architecture for the Internet of Things is proposed, with especial emphasis in its main module, the Service Orchestrator, for which implementation details are provided, including a model for the service placement task. Furthermore, technologies to implement the modules from the architecture are suggested. This proposal, as well as its validation, is framed within the scope of the SusCity project.