With the growing reliance on cloud computing for both personal and enterprise-level applications, the need for robust data security has become essential. Confidential Computing Environments (CCEs) offer a promising solution by providing hardware-based isolation to protect sensitive data and computations. It encrypts the main memory and creates an isolated execution environment for executing the workload. Among the leading CCE technologies, Intel's Trusted Domain Extensions (TDX) and AMD's Secure Encrypted Virtualisation with Secure Nested Paging (SEV-SNP) have emerged as significant solutions. However, it becomes essential to understand their performance impact across a range of applications. The challenge lies in balancing security and performance, especially for organisations deploying resource-intensive workloads in the cloud. This study is motivated by the need to offer a comprehensive, applicationlevel performance comparison of Intel TDX and AMD SEVSNP in a real-world cloud environment. By evaluating both CCEs for micro benchmarks and application microservices on the Microsoft Azure cloud environment, this paper aims to provide insights into how these technologies handle different types of workloads, thus enabling cloud users to make informed decisions based on their security and performance requirements.

Smart home systems, which integrate diverse IoT devices for automation and energy efficiency, are increasingly targeted by cyber threats such as Man-in-the-Middle (MitM) and Denial-of-Service (DoS)/Distributed DoS (DDoS) attacks. Traditional security mechanisms often lack the adaptability and responsiveness needed for real-time protection, leaving devices vulnerable to data breaches and manipulation. This paper proposes a behavioural fingerprinting-based security framework that combines threshold-based detection with machine learning (ML) classification to identify anomalies in device activity patterns, including round-trip time (RTT) and Address Resolution Protocol (ARP) behaviour. The proposed approach is evaluated in a real smart home testbed environment using an IP camera, Raspberry Pis, and an ASUS router. Experimental results demonstrate a high detection accuracy across varied attack scenarios. The results show that the proposed system can detect attacks in near real-time, while maintaining compatibility with third-party applications, making it a practical and effective solution for enhancing smart home security.

Ensuring Data Quality (DQ) is crucial for the reliability of environmental monitoring systems in the smart city environment. In this work, we present an automated framework, Data Quality Detector (DQD), for detecting, classifying, and analyzing DQ issues. We leveraged the Internet of Data Fault Taxonomy (IoDFT) to categorise faults based on various metrics, including type, duration, and pitfalls, providing a structured approach to identifying anomalies. By integrating these fault characteristics with key DQ dimensions-completeness, timeliness, and consis-tency- DQD enables a more granular classification of anomalies. Instead of simply marking missing faults, DQD differentiates between general and contextual missing faults, reducing false positives and improving decision-making accuracy. The DQD framework has been evaluated on real data from Newcastle Urban Observatory (NUO) to demonstrate its ability to enhance fault detection, minimise misclassification, and distinguish normal variations from actual quality issues, ultimately improving the trustworthiness and effectiveness of Air Quality (AQ) monitoring systems.

ABSTRACT Internet of Things systems exist in various areas of our everyday life. Data from sensors installed in smart cities and homes is processed in edge and cloud computing centers, providing several benefits that improve our lives. The group of devices might cooperate to fulfill desired goals, trying to preserve their resources and handle the failures of the devices. The paper presents the multi‐agent reinforcement learning (MARL) extension to the osmotic computing simulation framework, enabling direct interactions between IoT devices. The proposed approach allows IoT devices to operate autonomously and cooperatively in dynamic environments, reducing the need for manual intervention and enabling resilient, energy‐efficient sensing coverage. We discuss the multi‐agent sensing coverage problem as one directly applicable to the IoT sensing systems. We identify the challenges posed to the framework and analyze management algorithms for cooperating osmotic agents. In the evaluation, we demonstrate that cooperation between devices enables the self‐autonomous behavior of IoT systems. A case study yields promising results, showing that the adaptation of sensors may allow them to replace each other by lowering energy usage (up to 25% reduction), increasing global coverage (from 74% to 92%), and preserving battery life by dynamically adjusting their sensing range. Finally, the presented framework is a novel contribution that combines MARL environments with IoT systems simulation, enabling future research in this field.

As machine learning moves towards edge deployment, lifelong learning becomes crucial due to evolving data distributions and new tasks. Yet, applying traditional methods to learn from vast, complex IoT data streams poses challenges. These include excessive CPU usage, RAM overflow, prolonged convergence times disrupting device operation, and difficulties in adapting to concept drift. Consequently, models trained on devices struggle to handle frequently changing data, affecting their ability to respond effectively to new inputs.To address these issues, we introduce Poly Instance Lifelong Learning (PILL), an algorithm designed for real-time on-device model training and inference at the edge under lifelong learning settings. PILL is lightweight, operating efficiently on the CPUs of low-power single-board computers (SBCs). It achieves this by partitioning input data into manageable instances, filtering out label noise, and applying early stopping for rapid predictions.PILL was evaluated on three popular low-power SBCs as well as a high-end Windows 10 machine using four datasets of different sizes and features. The results indicate that despite the superior resources of the Windows 10 machine, models trained using PILL on SBCs differ in accuracy by only ±0.05%. Additionally, PILL’s LSTM trains 2.41 - 2.85 times faster than the widely used Scikit-Learn’s LSTM. Additionally, when compared to ten state-of-the-art methods, PILL demonstrated superior performance across key metrics (Precision, Recall, and F1-Score) while minimizing computational overhead, making it an ideal choice for efficient, real-time edge deployment.

We observe significant interest in reinforcement learning methods for real-world sensing-control scenarios driven by the sensor data streams. However, the delay introduced to the data by the communication channels may degrade the system's performance. It is especially crucial in the internet of things (IoT), where devices with constraint resources and low throughput networks are used.

Abstract Internet of Things systems exists in various areas of our everyday life. For example, sensors installed in smart cities and homes are processed in edge and cloud computing centers providing several benefits that improve our lives. The place of data processing is related to the required system response times—processing data closer to its source results in a shorter system response time. The osmotic computing concept enables flexible deployment of data processing services and their possible movement, just like particles in the osmosis phenomenon move between regions of different densities. At the same time, the impact of complex computer architecture on the environment is increasingly being compensated by the use of renewable and low‐carbon energy sources. However, the uncertainty of supplying green energy makes the management of osmotic computing demanding, and therefore their autonomy is desirable. In the article, we present a framework enabling osmotic computing simulation based on renewable energy sources and autonomic osmotic agents, allowing the analysis of distributed management algorithms. We discuss the challenges posed to the framework and analyze various management algorithms for cooperating osmotic agents. In the evaluation we show that changing the adaptation logic of the osmotic agents, it is possible to increase the self‐consumption of renewable energy sources or increase the usage of low emission ones.

Edge computing has gained momentum in recent years, as complementary to cloud computing, for supporting applications (e.g., industrial control systems) that require time-critical communication guarantees. While edge computing can provide immediate analysis of streaming data from Internet of Things devices, those devices lack computing capabilities to guarantee reasonable performance for time-critical applications. To alleviate this critical problem, the prevalent trend is to offload these data analytic tasks from the edge devices to the cloud. However, existing offloading approaches are static in nature as they are unable to adapt varying workload and network conditions. To handle these issues, we present a novel distributed and quality of services based multilevel queue traffic scheduling system that can undertake semiautomatic bandwidth slicing to process time-critical incoming traffic in the edge-cloud environments. Our developed system shows a great enhancement in latency and throughput as well as reduction in energy consumption for edge-cloud environments.

Along with an increase in the number of IoT devices and computations progressively shifting to the network edge and mobile systems, the complexity of computer systems is on the rise. In order to keep them in operation their monitoring becomes important. However, classic monitoring systems do not cope well with lightweight solutions due to the large volume of monitoring data sent to the cloud. In this paper we present mechanisms which can be applied to monitoring systems to make them suitable for IoT and mobile devices. We discuss and evaluate the benefits and drawbacks of both mechanisms as well as the results they produce. Finally, we propose the concept of function-driven log processor that can be used to control the proposed mechanisms.

This article presents a concept of a system which can be utilized as a remote management add-on for embedded devices.It can be applied to resource-constrained wireless sensors and IoT nodes based on a general purpose microcontroller unit or a field programmable gate array (FPGA) chip.The proposed solution facilitates remote firmware update, management, and operation monitoring.Thanks to the utilization of standard protocols and interfaces, the proposed system is very flexible and it can be easily customized for multiple modern microcontrollers or programmable logic chips.The presented system can be an efficient solution for fast prototyping and it can be an alternative to a time-consuming process of bootloader development for ad hoc devices.It can also be applied to remote laboratory access for educational purposes.A proof of concept prototype implementation has been successfully developed and evaluated.The implementation is available on a free license and utilizes a commonly available and inexpensive hardware platform.

In this paper we introduce the Rapid Embedded Systems Prototyping (RESP) approach aimed at accelerating the development of novel, experimental, and proof-of-concept implementations of embedded devices based on microcontrollers and Field Programmable Gate Array (FPGA) chips.It is intended to be used in the fast-paced business environment in which an early working prototype is required.The RESP approach can be useful for remote developing and temporary monitoring of various embedded devices: primarily for resource-constrained IoT platforms, microcontroller-based sensor nodes, and customized ad hoc systems.The RESP-compliant system uses a central server and one or multiple Remote Reconfiguration and Monitoring (RRM) modules.Each RRM allows the software developers to manage reprogramming and monitoring of multiple target embedded devices.It can be applied to a device that needs to be remotely reconfigured, tested, or reprogrammed in its target environment without implementing a reliable bootloader.The RRM described in this paper has been successfully implemented and its functionality and performance have been tested.

The Internet of Things concept has found applications in a wide range of solutions, including smart homes, smart cities, enterprise systems, and others. Due to the limited resources available on physical devices and the intermittent availability of wireless networks, IoT may hugely benefit from Mobile Cloud Computing as it may provide the necessary processing power and the storage. This, however, increases the complexity of IoT systems and calls for a flexible testing environment to simplify the development process. In this paper we propose the concept of a software‐defined IoT testbed which enables an IoT system to be immersed and tested in a virtual environment in order to evaluate its behavior under controllable conditions. Additionally, the features of the implemented testbed prototype are discussed on the basis of a specific use case.

The number of devices connected to the Internet is constantly growing. They are interacting with each other and produce data that has to be processed. The increasing data volume generated by the IoT devices is well handled by the flow based programming where data is moved through the networks of processes. Limited resources of the devices are compensated by the clouds creating the continuum. Nevertheless, to increase responsiveness in IoT or optimize the network bandwidth it might be necessary to move some processing to the fog i.e. to the devices located closer to the data sources. Execution of data flow on the number of heterogeneous IoT devices is not trivial as they offer different computational resources. In the paper, the concept of data flow transformation in order to execute parts of it, closer to the sources of data, on the devices with constrained resources is presented.

The paper proposes the concept of adaptive message aggregation for telemetry applications that use GPRS connectivity. The method optimizes the power consumed during data transmission, what is useful in the modern telemetry devices powered from renewable energy sources. The concept has been verified in the levee monitoring scenario.

This paper presents the concept of an Adaptive SOA Solution Stack (AS3). It is an extension of the S3 model, implemented via uniform application of the AS3 element pattern across different layers of the model. The pattern consists of components constituting an adaptation loop. The functionality of each component is specified in a generic way. Aspects of these patterns are analyzed in relation to individual S3 layers. The ability to achieve multilayer adaptation, provided by several cooperating AS3 elements is also discussed. Practical usage of the proposed concepts for Adaptive Operational Systems, Integration, and Service Component layers are presented in the form of three case studies. Each study describes the architecture of the proposed system extensions, selected software technologies, implementation details, and sample applications. Related work is discussed in order to provide a background for the reported research. This paper ends with conclusions and an outline of future work.

The adaptability of software applications is a very demanding feature nowadays. The main contribution of this paper is to propose a concept of policy-driven software framework supporting adaptable components called spontaneous objects. Spontaneous objects are lightweight adaptable context-aware software components. The adaptability mechanism is split between spontaneous objects and a spontaneous object container which creates a runtime environment for spontaneous objects. The paper describes the benefits achieved due to the policy-driven approach, a complete runtime architecture for spontaneous object implementation, a programming model and a case study