In this paper, we create a distribution of the embedded operating system for Cyclone V SoC FPGA platforms based on the Linux kernel. A comparison of well-known open-source tools for creating embedded operating systems is demonstrated. A step-by-step example of embedded OS synthesis using custom scripts and a simplified pipeline is given, which increases the adaptability of the target system. The possibility of adding hardware-oriented tools for interaction between SoC and FPGA is demonstrated. This makes it possible to create a wide range of hardware applications with remote access. The proposed approach is also vendor-independent and can be applied to other FPGA SoCs. The final system, although not significantly different in resource requirements from Yocto, is more adaptable and can be ported to the Yocto base if necessary. This allows us to make the most of the Full Custom approach, ensuring an optimal balance between development efficiency, responsiveness to changes, and system resource requirements.
11 December 2025 in Measuring and computing devices in technological processes
Робота розглядає автономний SDR-сканер для відтворюваних радіотехнічних експериментів у середовищах з обмеженими мережевими та обчислювальними ресурсами. Поставлена задача - забезпечити жорстку причинність та відтворюваність «від приймача до перегляду» без залежності від зовнішніх брокерів і TS-баз: кадр формується в C-вузлі (HackRF + FFTW, N=4096), серіалізується у NDJSON (у форматі «1 подія = 1 рядок») і передається в тонкий бекенд (FastAPI + SQLite/WAL) для одночасного запису та live-стріму до веб-клієнта (SSE/WS). Керування параметрами (Freq/SR/LNA/VGA/Atten) здійснюється через Unix Domain Socket із транзакційною парою txid↔ack, при чому команда застосовується між кадрами. Така організація зберігає часову геометрію (збереження кадрів спирається на часові мітки) й частотні інваріанти (bin_hz), спрощує трасування та дозволяє точне відтворення «як записано». Показано офлайн-систему на Raspberry Pi (ARM64), а також підхід до агрегування інтервалів спостережень (на основі часових проміжків) і керування життєвим циклом збережених проміжків. Запропонована архітектура мінімізує рухомі частини, але залишається здатною до масштабування, що робить її придатною для польових вимірювань, автономного моніторингу, прототипування методів радіоконтролю та розвитку власних рішень.
30 December 2025 in Security of Infocommunication Systems and Internet of Things
The article examines the role and future prospects of programmable logic devices (FPGAs) and system-on-chip FPGA (SoC FPGA) platforms within the edge computing paradigm. Particular attention is given to the combination of adaptability, fine-grained power consumption control, and high degrees of parallelism, which are critical characteristics for modern edge platforms. Additionally, the current state of FPGA adoption in practical edge scenarios is analyzed, ranging from video analytics in transportation systems to industrial vibration diagnostics and acceleration of telecommunications functions. Examples of both conventional FPGA-based solutions and hybrid SoC FPGA architectures are discussed, where programmable logic is tightly integrated with ARM-based processors to achieve balanced workload distribution between software and hardware components. It is demonstrated that such systems can significantly reduce processing latency, optimize energy consumption, and enable autonomous operation even under remote or unstable network conditions. The role of embedded operating systems is also examined, particularly in the context of SoC FPGA platforms, where Linux-based environments enable separation between control logic and hardware-accelerated data paths. The paper discusses how embedded operating systems influence system flexibility, software reuse, update mechanisms, and long-term maintainability of FPGA-based edge devices. In addition, the work addresses technical and organizational barriers that limit large-scale deployment of FPGA-based edge systems, including the lack of unified high-level synthesis toolchains, the steep learning curve associated with HDL-based design, and limited support for infrastructure-oriented workflows in FPGA-centric environments. Approaches for dynamic logic reconfiguration aimed at improving adaptability of local IoT systems are considered, along with challenges related to logic testing, system configuration, and scalability when adapting platforms to new application requirements. Modern development tools and frameworks for flexible system design, including cloud-based services and high-level programming environments such as Vitis HLS and Intel oneAPI, are also discussed in the context of reducing development complexity and accelerating design iterations.
24 September 2025 in Herald of Khmelnytskyi National University. Technical sciences
Article presents a modular telemetry platform for real-time collecting, processing, and presenting data that is relying on edge computing principles. The architecture integrates sensors as data producers, controllers (Raspberry PI and SoC) for processing data, and a web interface used to present and operate that data, into a unified architecture with less reliance on external networks. This method lets the platform function without reliance on global network therefore it can do it job in environments that have no reliable network coverage. Local data transfer is built around usage of MQTT broker, making sure that messages are sent and received, while keeping whole process asynchronous. All the parts function together in a localized Wi-Fi network, which makes it possible for data to be shared without needing additional transport laters. This setup does away with the requirement for cloud servers while still keeping data safe and making sure the system can keep running. SQLite, when set to write-ahead logging mode, is a strong way to save data locally and lets you quickly get back to work after an unforeseen error. The visualization layer, which is built on the Leaflet library, allows real-time map rendering and dynamic data updates through Server-Sent Events (SSE). This gives operators a clear, constantly updated view of the system's state. Connecting SoC FPGA modules and other devices through UART and SPI interfaces makes it possible to sync data at the hardware level and opens up new ways to scale the system. The architecture described provides a simple, modular, standalone solution. It meets the specifications for modern industrial IoT, so it can be used for research, field deployment, or laboratory automation. It is possible to easily add more sensors, data processing units, or communication interfaces to the system. Also, the whole approach allows possible extension to actually work with global network and distant compute power if needed in centralized structure arises. This makes it a flexible base for future development while keeping its essential functions and stability.
Continuous stream-oriented data-plane processing in 5G/6G edge nodes increases the energy and latency cost of CPU-centered execution. This paper studies this boundary on an Intel Cyclone V SoC FPGA and proposes Hardware Accountability: a partitioning discipline in which Linux performs supervisory control while high-rate payload processing remains in programmable logic. The evaluation uses the Strumok stream cipher, adopted as the Ukrainian national standard DSTU 8845:2019, as a secure fronthaul/payload workload with XOR- and shift-dominated logic. On the evaluated USB 2.0/Cortex-A9/Linux path, the software-driven stream approaches saturation near 20 MSPS. In contrast, the RTL core reaches 9.6 Gbps at 150 MHz and occupies less than 6% of the available logic. Quartus Prime vectorless power analysis estimates 24.00 mW dynamic power for the RTL computational core, corresponding to approximately 2.5 pJ/bit. Control-plane measurements show P99 orchestration jitter below 1 ms under Spatial Isolation, conservative full context reloads near 1290 per second, and more than 7200 shadow-register context/state update operations per second. A design-space exploration then projects an 83.2 Gbps multi-core data path when external DDR traffic is avoided through internal stream aggregation and elastic buffering.
30 December 2024 in Security of Infocommunication Systems and Internet of Things
Embedded systems and Edge Computing in Internet of Things (IoT) represent a special approach to creating systems for collecting, processing and analyzing data in an unstable environment. This article examines the benefits of Linux in this context, highlighting its flexibility, robust software ecosystems, and scalability, which are critical for a variety of IoT applications. The question of the operation of devices in an environment with a weak and/or unstable network is also raised, the general development of IoT/Internet-of-Everything (IoE) as a technology in the conditions of various distribution of high-speed networks is considered. Edge Computing technology, its use and areas of application in the need for rapid adaptation to the environment are also taken into account. Because the ability to calculate and analyze data on a local network can be critical for simplifying infrastructure in remote areas or in environments where access to an external network is difficult or impossible. We focus on the Linux kernel because its versatility in IoT is highlighted by its ability to handle a variety of workloads and seamlessly integrate with services, increasing adaptability to changing environmental conditions and ensuring reliable data processing at the edge. This adaptability is critical to mitigating the challenges caused by unreliable network infrastructure, thereby facilitating real-time decision-making and increasing operational efficiency. In addition, the open nature of Linux fosters innovation, allowing developers to create solutions tailored to the specific needs of edge computing, from industrial automation to smart city initiatives. By allowing devices to operate autonomously and efficiently manage resources at the network edge, Linux significantly optimizes latency, resource utilization, and overall system performance. Use of edge computing with correctly set-up embedded operating system (OS) allows to avoid issues common in IoT field and related to environment change. Article provides insight into pros and cons edge computing, its implementation in IoT and IoE by embedded Linux based OS. We will go through most common use-cases and market shares of common OS options. While IoT takes part in most industries by storm, there are still problems common for new industry. The primary advantage of using embedded *nix OS is the agility and ease of incorporation of those devices into edge computing systems, allowing to deal with network issues. Due to IoT/IoE being a new industry where many technologies are combined there are a lot of different approaches and frameworks that are used in it, but some of them are more popular and common than other ones. While going through the IoT/IoE data in the article, we will focus on embedded edge computing as one of the most efficient ways in building IoT solutions. Especially in perspective of OS market changes now and in near future. The result of that study will provide insight into possible trends and positives of the use of embedded OS with edge computing.