The rapid proliferation of Internet of Things (IoT) technologies has transformed the computing landscape, placing unprecedented demands on traditional cloud-centric architectures. The massive volume of distributed data, coupled with stringent latency and reliability requirements, has accelerated the adoption of edge-oriented computing paradigms. Within this context, Fog and Mist computing have become fundamental components of the edge computing system, enabling computation to be progressively shifted from centralized cloud platforms toward network edges and even directly onto end devices. This keynote examines the challenges and opportunities associated with orchestrating computational resources across the cloud-fog-mist continuum, with particular emphasis on scheduling mechanisms for real-time and missioncritical IoT applications. It reviews contemporary approaches that balance latency, resource utilization, energy efficiency, and quality of service while satisfying application deadlines in highly dynamic environments. The presentation also discusses emerging technologies, including adaptive scheduling, highlighting the key research challenges that will shape next-generation edge computing infrastructures.
The rapid expansion of the Internet of Things (IoT) has created a heterogeneous attack surface that spans consumer devices, enterprise systems, industrial control networks, and critical infrastructure. Although prior work has examined vulnerabilities within individual domains, there is limited empirical evidence comparing security exposure across operational sectors. This paper presents a cross-sector measurement study of publicly reachable IoT systems in consumer, commercial, industrial, infrastructure, and military environments. Using indexed Internet discovery and controlled non-intrusive vulnerability assessment, we constructed a dataset comprising 424 unique vulnerabilities and 12,366 vulnerability instances across multiple device classes. The analysis evaluates severity distribution, vulnerability categories, and sector-specific exposure behavior, showing that IoT insecurity is not uniform: consumer and commercial environments exhibit high-frequency web and configuration weaknesses, while industrial and infrastructure systems demonstrate fewer but higher-impact vulnerabilities related to control protocols and segmentation. To improve risk prioritization, an Isolation Forest-based anomaly detection model was applied to identify high-risk IoT exposure profiles based on vulnerability density and severity characteristics. The results indicate that IoT risk is primarily shaped by deployment context, highlighting the need for sector-aware defensive strategies and scalable AI-assisted analysis approaches.
The growing deployment of Internet of Things (IoT) devices has intensified the demand for energy-efficient communication systems that ensure reliable operation under stringent power and maintenance constraints. Traditional transceiver designs often struggle to achieve long-term energy sustainability, resulting in frequent battery replacements, high operational costs, and service disruptions, particularly in remote or difficult-toaccess IoT deployments. This paper presents a novel energyefficient transceiver architecture tailored for IoT applications, integrating hybrid modulation with intelligent power amplifier (PA) control to reduce energy consumption. By selectively deactivating the PA for statistically dominant “typical” symbols, the proposed system achieves up to 30% energy savings without sacrificing communication quality. To support efficient signal decoding, two receiver architectures are introduced: (i) an energy detectionbased approach with a closed-form Bit Error Rate (BER) model, and (ii) a deep learning-based approach leveraging denoising and classification for enhanced robustness. Simulation results confirm that the proposed design maintains BER performance comparable to conventional schemes while preserving spectral efficiency and significantly extending battery life. This solution offers a scalable, low-power communication method suitable for IoT deployments that require reliable and long-duration operation.
Embedded systems present a set of requirements to grant a secure and efficient use of their resources, like reliable communication, wireless connectivity, low power consumption, and safe data management. This paper explores suitable file system solutions for embedded systems, focusing on three prominent options: FatFS, LittleFS, and Spiffs. Firstly, a study of existing solutions is presented. Secondly, the main aspects of the architectures are analyzed by categorizing them on work principles. Then, a study methodology is proposed to evaluate the file systems using throughput and latency criteria. Tests related to file opening, closing, reading, and writing operations were performed. The efficiency of in-memory data access is analyzed by selecting a set of files with different sizes. Lastly, the results were compared and discussed across the three file systems. The results show that FatFS is an option for faster file creation, presenting a 99.81% improvement over Spiffs and faster file reading with an average of $\mathbf{1 1. 4 6 ~ M B} / \mathrm{s}$. Spiffs is suitable for operations involving quick file opening (99.33% faster than LittleFS), file closing (99.97% faster than FatFS when closing a new file), and file writing with an average of $\mathbf{0. 1 2 ~ M B / s}$. For the power loss scenario, Spiffs is recommended since it was able to recover data in the event of unexpected power failure during write operations, reducing energy waste and improving system reliability in energy-constrained IoT applications.
This paper conducts a comprehensive evaluation of general-purpose compute instances offered by leading cloud service providers, focusing on the interplay between processor architecture, cost, and performance metrics. Utilizing standardized configurations and benchmarking methodologies across Intel, AMD, and ARM architectures, this research delineates the tradeoffs inherent in cloud infrastructure selection. Results demonstrate ARM-based instances provide superior cost-efficiency for scale-out and cloud-native workloads, while Intel architectures maintain dominance in legacy-sensitive, performance-critical environments. Insights derived aim to empower computing professionals in optimizing cloud resource allocation, maximizing computational throughput per expense, and guiding strategic deployments within evolving distributed systems paradigms.
