The increasing complexity of IoT applications and the continuous growth in data generated by connected devices have led to significant challenges in managing resources and meeting performance requirements in computing continuum architectures. Traditional cloud solutions struggle to handle the dynamic nature of these environments, where both infrastructure demands and data analytics requirements can fluctuate rapidly. As a result, there is a need for more adaptable and intelligent resource management solutions that can respond to these changes in real-time. This paper introduces a framework based on multi-dimensional elasticity, which enables the adaptive management of both infrastructure resources and data analytics requirements. The framework leverages an orchestrator capable of dynamically adjusting architecture resources such as CPU, memory, or bandwidth and modulating data analytics requirements, including coverage, sample, and freshness. The framework has been evaluated, demonstrating the impact of varying data analytics requirements on system performance and the orchestrator's effectiveness in maintaining a balanced and optimized system, ensuring efficient operation across edge and head nodes.

Machine learning (ML) typically relies on the assumption that training and testing distributions are identical and that data are centrally stored for training and testing. However, in real-world scenarios, distributions may differ significantly, and data are often distributed across different devices, organizations, or edge nodes. Consequently, it is to develop models capable of effectively generalizing across unseen distributions in data spanning various domains. In response to this challenge, there has been a surge of interest in federated domain generalization (FDG) in recent years. FDG synergizes federated learning (FL) and domain generalization (DG) techniques, facilitating collaborative model development across diverse source domains for effective generalization to unseen domains, all while maintaining data privacy. However, generalizing the federated model under domain shifts remains a complex, underexplored issue. This article provides a comprehensive survey of the latest advancements in this field. Initially, we discuss the development process from traditional ML to domain adaptation (DA) and DG, leading to FDG, as well as provide the corresponding formal definition. Subsequently, we classify recent methodologies into four distinct categories: federated domain alignment (FDAL), data manipulation (DM), learning strategies (LSs), and aggregation optimization (AO), detailing appropriate algorithms for each. We then overview commonly utilized datasets, applications, evaluations, and benchmarks. Conclusively, this survey outlines potential future research directions.

As a driving force in the advancement of intel-ligent in-orbit applications, DNN models have been gradually integrated into satellites, producing daily latency-constraint and computation-intensive tasks. However, the substantial computation capability of DNN models, coupled with the instability of the satellite-ground link, pose significant challenges, hindering the timely completion of tasks. It becomes necessary to adapt to task stream changes when dealing with tasks requiring latency guarantees, such as dynamic observation tasks on the satellites. To this end, we consider a system model for a collaborative inference system with latency constraints, leveraging the multi-exit and model partition technology. To address this, we propose an algorithm, which is tailored to effectively address the trade-off between task completion and maintaining satisfactory task accuracy by dynamically choosing early-exit and partition points. Simulation evaluations show that our proposed algorithm signif-icantly outperforms baseline algorithms across the task stream with strict latency constraints.

Hierarchical federated learning (HFL) designs introduce intermediate aggregator nodes between clients and the global federated learning server in order to reduce communication costs and distribute server load. One side effect is that machine learning model replication at scale comes “for free” as part of the HFL process: model replicas are hosted at the client end, intermediate nodes, and the global server level and are readily available for serving inference requests. This creates opportunities for efficient model serving but simultaneously couples the training and serving processes and calls for their joint orchestration. This is particularly important for continual learning, where serving a model while (re)training it periodically, upon specific triggers, or continuously, takes place over shared infrastructure spanning the computing continuum. Consequently, training and inference workloads can interfere with detrimental effects on performance. To address this issue, we propose an inference load-aware HFL orchestration scheme, which makes informed decisions on HFL configuration, considering knowledge about inference workloads and the respective processing capacity. Applying our scheme to a continual learning use case in the transportation domain, we demonstrate that by optimizing aggregator node placement and device-aggregator association, significant inference latency savings can be achieved while communication costs are drastically reduced compared to flat centralized federated learning.

The growing number of Internet of Things (IoT) devices generates massive amounts of diverse data, including personal or confidential information (i.e., sensory, images, etc.) that is not intended for public view. Traditionally, predefined privacy policies are usually enforced in resource-rich environments such as the cloud to protect sensitive information from being released. However, the massive amount of data streams, heterogeneous devices, and networks involved affects latency, and the possibility of having data intercepted grows as it travels away from the data source. Therefore, such data streams must be transformed on the IoT device or within available devices (i.e., edge devices) in its vicinity to ensure privacy. In this paper, we present a privacy-enforcing framework that transforms data streams on edge networks. We treat privacy close to the data source, using powerful edge devices to perform various operations to ensure privacy. Whenever an IoT device captures personal or confidential data, an edge gateway in the device’s vicinity analyzes and transforms data streams according to a predefined set of rules. How and when data is modified is defined precisely by a set of triggers and transformations - a privacy model - that directly represents a stakeholder’s privacy policies. Our work answered how to represent such privacy policies in a model and enforce transformations on the edge.

Recent advancements in distributed systems have enabled deploying low-latency and highly resilient edge applications close to the IoT domain at the edge of the network. The broad range of edge application requirements combined with heterogeneous, resource-constrained, and dynamic edge networks make it particularly challenging to configure and deploy them. Besides that, missing elastic capabilities on the edge makes it difficult to operate such applications under dynamic workloads. To this end, this article proposes a lightweight, self-adaptive, and decentralized mechanism (DECENT) for (1) deploying edge applications on edge resources and on premises of Edge-Cloud infrastructure and (2) controlling elasticity requirements. DECENT enables developers to characterize their edge applications by specifying elasticity requirements, which are automatically captured, interpreted, and enforced by our decentralized elasticity interpreters. In response to dynamic workloads, edge applications automatically adapt in compliance with their elasticity requirements. We discuss the architecture, processes of the approach, and the experiment conducted on a real-world testbed to validate its feasibility on low-powered edge devices. Furthermore, we show performance and adaptation aspects through an edge safety application and its evolution in elasticity space (i.e., cost, resource, and quality).

Recent advancements in distributed systems have enabled deploying low-latency edge applications (i.e., IoT applications) in proximity to the end-users, respectively, in edge networks. The stringent requirements combined with heterogeneous, resource-constrained and dynamic edge networks make the deployment process a challenging task. Besides that, the lack of resource discovery features make it particularly difficult to fully exploit available resources (i.e., computational, storage, and IoT resources) provided by low-powered edge devices. To that end, this article proposes a decentralized resource discovery mechanism that enables discovering resources in an automatic manner in edge networks. Through replicating resource descriptions (i.e., metadata), edge devices exchange information about available resources within their scope in a peer-to-peer manner. To handle the resource discovery complexity, we propose a solution to built edge networks as a flat model and enable edge devices to be organized in clusters. Our approach supports the system in coping with the dynamicity and uncertainty of edge networks. We discuss the architecture, processes of the approach, and the experiments we conducted on a testbed to validate its feasibility on resource-constrained edge networks.

Engineering Internet of Things (IoT) systems is a challenging task partly due to the dynamicity and uncertainty of the environment including the involvement of the human in the loop. Users should be able to achieve their goals seamlessly in different environments, and IoT systems should be able to cope with dynamic changes. Several approaches have been proposed to enable the automated formation, enactment, and self-adaptation of goal-driven IoT systems. However, they do not address deployment issues. In this paper, we propose a goal-driven approach for deploying self-adaptive IoT systems in the Edge-Cloud continuum. Our approach supports the systems to cope with the dynamicity and uncertainty of the environment including changes in their deployment topologies, i.e., the deployment nodes and their interconnections. We describe the architecture and processes of the approach and the simulations that we conducted to validate its feasibility. The results of the simulations show that the approach scales well when generating and adapting the deployment topologies of goal-driven IoT systems in smart homes and smart buildings.

Edge computing has been recently introduced as an intermediary between Internet of Things (IoT) deployments and the cloud, providing data or control facilities to participating IoT devices. This includes actively supporting IoT resource discovery, something particularly pertinent when building large-scale, distributed and heterogeneous IoT systems. Moreover, edge devices supporting resource discovery are required to meet the stringent requirements prevalent in IoT systems including high availability, low-latency, and privacy. To this end, we present a resource discovery platform for IoT resources situated at the edge of the network. Our approach aims at providing a seamless discovery process that is able to (i) extend the covered area by deploying additional edge nodes and (ii) assist in the development of new IoT applications that target already available resources. Within our proposed platform, devices located in a certain proximity connect and form an edge-to-edge network that we call an edge neighborhood - our edge-to-edge metadata replication platform enables participating devices to discover available resources. Our solution is characterized by absence of centralization, as edge nodes exchange metadata about available resources within their scope in a peer-to-peer manner.

Software components within heterogeneous devices of the Internet of Things (IoT) systems use resources representing various computational capabilities, including sensing or actuation end points. However, components do not live in isolation and must be able to coordinate with others to fulfill their goals. Satisfaction of requirements-capturing their goals-must persist in environments that are changing, unpredictable, and potentially unknown at system design time. Edge computers placed near IoT devices can be leveraged for this sort of control-providing resource management for end devices within their operational context. We propose a methodology and technical framework for engineering resource coordination at runtime, tailored for the decentralized, pervasive systems of today. Our approach represents a paradigm shift in marrying distributed systems and formal aspects of software engineering. We adopt goal modeling to capture objectives within the system and use bounded model checking as the foundational technique to compute coordination plans that satisfy device goals. This occurs opportunistically at runtime without any knowledge about the operational status or presence of resources, but always in accordance with the edge's own goals. Our technical framework exhibits dependability guarantees regarding optimality and correctness of generated plans. We evaluate the resource coordination performance and its feasibility on low-powered ARM-based edge devices.