Internet of Things (IoT) edge nodes are often battery-powered or energy-harvesting and must operate under tight energy and area constraints while still supporting low-latency local decision-making. Since wireless transmission is energy-expensive, IoT increasingly relies on in-situ edge processing, which demands ultra-low-power and compact compute blocks, yet further complementary-metal-oxide-semiconductor (CMOS) scaling is constrained by leakage, interconnect congestion, and rising power density. Atomic silicon quantum-dot (ASQD) technology offers a promising alternative by enabling computation via charge configuration rather than current flow, potentially providing very low energy operation and highly compact layouts; however, large-scale ASQD circuits remain challenging due to coupling-induced stability limitations. To address this, we propose a low-power ASQD-based arithmetic circuit for low-power edge IoT computing that supports four distinct operations under a unified multiplexer (MUX) controlled structure. The design follows an integration-aware methodology in which candidate primitives are benchmarked at the gate level (Atoms/Area/Energy), and optimized variants are selected to minimize cumulative arithmetic logic unit (ALU) level cost. ALU-level scaling shows reductions of approximately 69–79% in occupied area and 43–51% in total energy compared with representative literature libraries, and SiQAD simulations verify stable operation. These results indicate that the proposed circuit can serve as an enabling compute primitive for future ASQD IoT platforms.

The rapid expansion of Internet of Things (IoT) devices has introduced a new phase of inter connectivity and convenience, while also presenting notable security obstacles. This research paper investigates novel methodologies for enhancing the security of IoT networks by using machine learning and deep learning methodologies. In order to accomplish this objective, two well acknowledged datasets, namely UNSW-NB15 and BoT-IoT, are used for the purpose of constructing and assessing resilient security models. The UNSW-NB15 dataset is well recognized for its extensive compilation of network traffic data, while the BoT-IoT dataset is especially designed to facilitate study on security issues related to the IoT. These datasets provide a wide array of attack scenarios and network traffic patterns, therefore enhancing the diversity of available resources for researchers in this field. By using these datasets, our research endeavours to examine many crucial facets of security inside IoT networks. Initially, machine learning techniques are used for the purpose of conducting intrusion detection on IoT networks. This stage encompasses the categorization of network traffic into two distinct categories: normal and malicious. This process facilitates the prompt detection and reaction to potential threats in real-time. Deep learning methodologies demonstrate exceptional proficiency in capturing complicated patterns and behaviours inherent in IoT network traffic, hence enhancing the capacity to detect intricate and dynamic security risks. In order to assure the practical usability of the presented models, we take into account variables such as computing efficiency and scalability. This is particularly important since IoT devices often have limited resources available. Furthermore, we investigate approaches aimed at enhancing the interpretability of models, therefore offering valuable insights into the underlying decision-making mechanisms, ultimately fostering trust and transparency.

Outsourced software development using the DevOps Model (COSD) has become very popular in the tech industry due to its financial and strategic benefits. The DevOps (COSD) Model facilitates the development process of software across different geographical locations for developer virtual. However, there are many difficulties in managing and organizing the development processes/activities due to cultural and time zone differences in distributed geographical locations. In this research project, our aim is to develop a readiness model that could improve the DevOps Model (COSD) for managing software development activities through assessment using the readiness model. We will compare the structural design of our model with the Capability Maturity Model Integration (CMMI). Additionally, we will consider the most critical factors in terms of success and barriers as readiness levels highlighted in real-world industry practices and existing literature. Our proposed model provides an idea and emphasizes the importance of a maturity model for software development and successful management of COSD activities in the tech industry.

In the era of 6G, developing and managing software requires cutting-edge software engineering (SE) theories and practices tailored for such complexity across a vast number of connected edge devices. Our project aims to lead the development of sustainable methods and energy-efficient orchestration models specifically for edge environments, enhancing architectural support driven by AI for contemporary edge-to-cloud continuum computing. This initiative seeks to position Finland at the forefront of the 6G landscape, focusing on sophisticated edge orchestration and robust software architectures to optimize the performance and scalability of edge networks. Collaborating with leading Finnish universities and companies, the project emphasizes deep industry-academia collaboration and international expertise to address critical challenges in edge orchestration and software architecture, aiming to drive significant advancements in software productivity and market impact.

Space-borne remote sensing provides great potential for soil moisture (SM) retrieval and emerged as a significant data source for research in land surface dynamics and associated applications. This study compared the error characteristics of SM estimates retrieved from the Advanced Microwave Scanning Radiometer 2 (AMSR2) instrument on board NASA’s Aqua satellite across different vegetation gradients and land cover conditions, at different overpass times and frequencies, to demonstrate their strengths and limitations. Results demonstrate that AMSR2 C-band products outperform AMSR2 X-band products over moderately and densely vegetated conditions due to lower attenuation by the vegetation canopy. Conversely, X-band products performed better than C-band products in barren lands possibly due to uneven sensing depth and microwave emissions from subsurface in C-band. The daytime products have a higher signal-to-noise ratio (SNR) in sparsely and moderately vegetated areas, whereas nighttime products have a higher SNR in densely vegetated areas. When these products are used selectively based on their error characteristics, the probability of obtaining SM with stronger signal than noise can be significantly improved (95%) at the expense of impaired spatial coverage (70%-pixel loss).