Hardware Trojans are emerging malicious integrated circuit (IC) modifications that pose a significant threat to the integrity of electronics. While existing methods, such as functional testing and reverse engineering, are proposed to identify Trojan anomalies in electronics, their applicability to industrial pipelines is limited. This paper proposes a new image processing technique for efficient clustering and identification of Hardware Trojan insertion in integrated circuits. The uniqueness of the proposed AI-assisted image processing method relies on using real hardware to generate images using side-channel analysis (SCA) before applying unsupervised image classification to identify the impact of hardware Trojans without the need for costly golden references. Leveraging Machine Learning on side-channel data collected from Ring-Oscillator networks, image and digital signal processing are employed to extract features for detection. This research contributes a novel use of side-channel data as images, eliminating the reliance on golden references, and achieving a remarkable accuracy of 95% in Hardware Trojan detection. In addition to significantly advancing the field and addressing crucial challenges in semiconductor supply chain, making it a significant step toward securing it.

Compromising in cyber security, especially regarding data leaks and single points of failure, pose a problem to the rapid growth of the Internet of Things (IoT). A solution for these issues is Blockchain and Federated Learning (FL) integration, which offer a more private and decentralized method for ensuring the security of IoT devices. This integration improves the integrity of data autonomously, while authentication and secure model aggregation block any attempts to tamper. Moreover, FL allows Artificial Intelligence (AI) to be trained on devices, making the exposure of data less likely. In this paper, we analyse the attack surface and emerging vulnerabilities of Blockchain-FL based IoT systems while constructing a taxonomy of the threats posed. They focus on novel security frameworks aimed at constructing next generation IoT system infrastructure, which is scalable, robust, intelligent and above all, secure.

Falls among older adults can cause serious injury and loss of independence. cStick 2.0 is a vision-enabled, IoMT-edge based smart walking-stick prototype that combines multimodal fall-risk classification with embedded obstacle awareness. The fall-risk classifiers were evaluated using a 9670-record development dataset, on which the compact DNN achieved 95.40% accuracy, 92.35% balanced accuracy, a macro F1-score of 93.90%, and a ROC-AUC of 97.44%. The Arduino Nicla Vision obstacle module used an INT8 Edge Impulse model with centroid-based direction assignment and time-of-flight distance sensing; 144 controlled trials produced 75.00% obstacle-presence accuracy at approximately 19–20 FPS. Sensor acquisition, GPS, display output, buzzer response, and CSV record accumulation were demonstrated at a prototype level. Synchronized older-adult evaluation, device-to-application communication, secure caregiver services, multimodal accessibility feedback, and longitudinal personalization remain future validation stages.

The Internet of Things (IoT) was introduced almost two decades ago. In the past two decades, technology has seen huge advancements. Many devices have become powerful and have less power consumption. Many IoT architectures and environments were introduced to help make life easier, especially in wearable devices. The market for these wearable devices has constantly increased over the years and is expected to reach its maximum in the next couple of years. They also pose a threat to users' privacy and security because they constantly store and transmit personal information such as location, heart rate, and other sensitive data. Therefore, addressing the security vulnerabilities is a crucial aspect of this research. This paper presents a hardware-assisted, energy-efficient, low-overhead security solution for wearable devices. Specifically, two Physical Unclonable Function (PUF) architectures: Arbiter PUF and Hybrid Oscillator Arbiter (HOA) PUF are analyzed for integration in IoT systems. The result shows that Arbiter PUF consumes 25 μW, whereas HOA PUF consumes only 2.7 μW to generate keys for cryptographic purposes. These architectures introduce minimal power overhead while providing robust security, making them well suited for resource-constrained IoT ecosystems.

Federated Learning (FL) enables collaborative training without centralizing sensitive data but faces challenges, including client authenticity, verifiable training participation, and secure aggregation. To overcome these challenges, we propose a novel framework, Zero-Knowledge Reputation-aware Blockchain Federated Learning (ZK-RBFL), which integrates blockchain, FL, Homomorphic Encryption (HE), and zero-knowledge proofs (ZKP). In the proposed ZK-RBFL framework, initially the clients undergo lightweight token-based authentication and then generate ZKP to provide cryptographic evidence of honest local training participation and reported inference accuracy before contributing their model updates. The model updates are encrypted using the CKKS HE mechanism to prevent any potential model inversion attacks. These encrypted model updates are stored on IPFS, with their corresponding CIDs recorded on the blockchain to ensure immutability. Further, ZK-RBFL enables mutual client verification of ZKPs to reduce server bottlenecks and enhance accountability. To ensure fairness and robustness in a distributed environment, we introduce a democratic blockchain consensus mechanism named Proof of Reputation-Weighted Voting (PoRWV) for block acceptance. Once consensus is reached, the encrypted model updates are aggregated using reputation-weighted averaging. We demonstrate the effectiveness of ZK-RBFL for brain tumor classification using a ZKP-compatible LeNet model for proof generation. Despite model simplicity, the global model achieves 94.22% accuracy. In addition, experiments with malicious clients and formal Scyther security analysis demonstrate that ZK-RBFL ensures both security and performance.

Effective diet management is crucial for preventing non-communicable diseases, including heart disease, diabetes, and obesity. This paper introduces NutriVision, an enhanced system featuring an interactive chatbot that delivers real-time dietary guidance and personalized recommendations. NutriVision enables users to inquire about nutritional content, receive customized recipe suggestions, and track nutritional goals through natural language interaction. By integrating user data such as health conditions, dietary preferences, and nutritional requirements, it generates tailored responses that enhance user engagement and decision-making. With the help of computer vision and machine learning, NutriVision accurately identifies food items and estimates quantities from smartphone-captured images, providing instant nutritional analysis that includes both macronutrient and micronutrient details with a 94

Agriculture 4.0 is usually presented through its most visible products: autonomous tractors, weed-spotting drones, and language-driven advisory apps. Field deployment has proven harder than these demonstrations suggest. Models that report accuracies above 99 % on curated leaf images frequently lose 25-40 percentage points when rerun on smartphone photographs from a working farm; reinforcement-learning irrigation controllers that save 30 % water in simulation routinely over-or underwater clay and sand soils that violate the simulator's homogeneity assumptions; and the machine-learning stack actually deployed across hundreds of millions of hectares remains dominated by gradient-boosted trees rather than the transformer architectures dominating the literature. This survey is organised around this gap between reported and deployed performance. We review the four agricultural revolutions, describe the hardware, connectivity, data, and intelligence layers of the modern stack, and map eleven model families onto eight application domains that cover most field-level decisions. Two analytical sections address deployment directly: one compares model families on the criteria that govern model choice in practice (latency, edge-feasibility, data efficiency, interpretability), and one documents the lab-to-field failure modes that benchmark scores hide. We close with seven research frontiers (geometric deep learning, agricultural foundation models, digital twins, agentic systems, continual learning, edge AI, federated data) and seven deployment barriers (domain shift, calibration, interpretability, smallholder cost, regulation, sovereignty, robustness). The survey is aimed at two readers: researchers entering the area who want a reliable map of what works, and practitioners who need to know where reported numbers should be treated with caution.

This research investigates the integration of quantum hardware-assisted security into critical applications, including the Industrial Internet-of-Things (IIoT), Smart Grid, and Smart Transportation. The Quantum Physical Unclonable Functions (QPUF) architecture has emerged as a robust security paradigm, harnessing the inherent randomness of quantum hardware to generate unique and tamper-resistant cryptographic fingerprints. This work explores the potential of Quantum Computing for Security-by-Design (SbD) in the Industrial Internet-of-Things (IIoT), aiming to establish security as a fundamental and inherent feature. SbD in Quantum Computing focuses on ensuring the security and privacy of Quantum computing applications by leveraging the fundamental principles of quantum mechanics, which underpin the quantum computing infrastructure. This research presents a scalable and sustainable security framework for the trusted attestation of smart industrial entities in Quantum Industrial Internet-of-Things (QIoT) applications within Industry 4.0. Central to this approach is the QPUF, which leverages quantum mechanical principles to generate unique, tamper-resistant fingerprints. The proposed QPUF circuit logic has been deployed on IBM quantum systems and simulators for validation. The experimental results demonstrate the enhanced randomness and an intra-hamming distance of approximately 50% on the IBM quantum hardware, along with improved reliability despite varying error rates, coherence, and decoherence times. Furthermore, the circuit achieved 100% reliability on Google’s Cirq simulator and 95% reliability on IBM’s quantum simulator, highlighting the QPUF’s potential in advancing quantum-centric security solutions.

The recent advancements in blockchain technology have also expanded its applications to smart agricultural fields, leading to increased research and studies in areas such as supply chain traceability systems and insurance systems. Policies and reward systems built on top of centralized systems face several problems and issues, including data integrity issues, modifications in data readings, third-party banking vulnerabilities, and central point failures. The current paper discusses how farming is becoming a leading cause of water and electricity wastage and introduces a novel idea called IncentiveChain. To keep a limit on the usage of resources in farming, we implemented an application for distributing cryptocurrency to the producers, as the farmers are responsible for the activities in farming fields. Launching incentive schemes can benefit farmers economically and attract more interest and attention. We provide a state-of-the-art architecture and design through distributed storage, which will include using edge points and various technologies affiliated with national agricultural departments and regional utility companies to make IncentiveChain practical. We successfully demonstrate the execution of the IncentiveChain application by transferring crypto-ether from utility company accounts to farmer accounts in a decentralized system application. With this system, the ether is distributed to the farmer more securely using the blockchain, which in turn removes third-party banking vulnerabilities and central, cloud, and blockchain constraints and adds data trust and authenticity.

Fault Injection attack is a type of side-channel attack on the Physical Unclonable Function (PUF) module that can induce faults in the PUF response by manipulating the PUF circuit behavior through voltage glitches, laser attacks, temperature manipulations, or any other attacks potentially leading to information loss or security system failure. This type of attack exposes the physical characteristics of PUFs that can be analyzed to predict or compromise the unique challenge-response pairs (CRPs) reducing the security and reliability of the PUF. Mitigation strategies against such attacks typically include adding noise to the PUF output, using error-correcting codes, or enhanced cryptographic protocols that obscure physical side-channel attacks. In this research, we propose a Generative Adversarial Network (GAN) based security model, that monitors the PUF behavior and detects the variations in PUF response. The model can detect glitches in the PUF response and generate alerts to take mitigation measures.

The Smart Grid concept evolved from the idea of intelligent and secure management of electrical grid infrastructure components and their communication through sustainable integration with the state-of-the-art technologies. This research focuses on emerging quantum computing-assisted security and its application in the Smart Grid. The robustness of electrical grid is increasing every day through advancements in grid infrastructure management which include outage control, relay protection, reliable distribution, renewable energy resource integration, and energy trading. Quantum Computing emerges as a formidable solution for application in the smart grid due to its processing capability and scale. Its application and scope are evolving every day with the recent developments in Quantum Chips which could pave the way for emerging Quantum-Chain-of-Things (QCoT). This research focuses on providing robust security in smart grids through Quantum Physical Unclonable Functions (QPUF) primitive, a quantum-hardware assisted security approach driven by micro manufacturing quantum process variations for generating a quantum digital fingerprint driven by quantum mechanics principles. The QPUF experimental evaluation in this research was performed to uniquely fingerprint various electrical grid entities providing a sustainable and secure flow of communication. Experimental evaluation shows a robust and reliable extraction of quantum digital fingerprints from noisy IBM quantum systems. The evaluation shows an impressive 86% keys achieving 100% reliability.

With the rapid advance in Deep Neural Networks (DNNs), GPU's role as a hardware accelerator becomes increasingly important. Due to the GPU's significant power consumption, developing highperformance and power-efficient GPU systems is a critical challenge. DNN applications need to move a large amount of data between memory and the processing cores, which consumes a great amount of NoC power. However, prior proposed lossless data compressions cannot achieve optimal performance and energy efficiency because they did not take advantage of the error resilience of DNNs. In this work, we propose an NoC architecture that can reduce power consumption without compromising performance and accuracy. Our technique takes advantage of the error resilience of DNNs as well as the data locality in the floating-point data representation of DNNs. Each data packet is reorganized by grouping data with similar bits such as in the exponents, and redundant bits are sent only once. We further compress the mantissa fields by appropriately selecting "proxy" values for data sharing the same exponent. Our evaluation results show that the proposed technique can effectively reduce the amount of data transmitted and lead to better performance and power trade-offs while preserving accuracy.

Street lighting is one of the prominent applications that demand a massive amount of power and substantially contributes to the energy budget of a country. Light Emitting Diode (LED) and the advancement of Internet of Things (IoT) have significantly improved conventional street light technology. Nevertheless, the rapid growth of IoT devices has presented a formidable challenge in powering the vast array of IoT devices. In this manuscript, a sustainable, battery-free, low-power street light management system has been proposed which is powered from hybrid solar and solar thermal energy harvesting scheme integrated with an efficient power management unit. As a specific case study, the prototype has been implemented with an existing LED street light in India. The characteristics and performance of the prototype have been evaluated to ensure its seamless operation under real-world scenarios. The average power consumption of the system is measured as 2.088 mW when operating in real-time with 50% duty cycle, exhibiting high Quality of Service (QoS). It features long-range communication up to 761 m through implementing LoRaWAN technology. Dimension of the prototype has been restricted to 10.5 cm x 6.5 cm x 2.3 cm to make it suitable for retrofitting with existing LED based street lights

Federated learning (FL) is a valuable solution for training models on distributed data while maintaining privacy. However, FL also introduces new security threats such as, poisoning attacks. While studies have addressed FL security, they often rely on centralized defenses, and the natural diversity of data is not considered, which limits their generalization in real-world applications. In this paper, we propose FedSecure, a decentralized and adaptive anomaly detection framework to mitigate poisoning attacks in the Internet of Medical Things (IoMT). Our proposal integrates a Bi-LSTM autoencoder and DNN models for anomaly detection. FedSecure was evaluated on real-world datasets, achieving 95.4% accuracy with Bi-LSTM and $99.96 \%$ with DNN. Experimental results demonstrate its effectiveness in enhancing FL security against poisoning attacks in IoMT environments.

The rise of the Internet of Medical Things (IoMT) enables continuous health monitoring, but traditional cloud-based Machine Learning (ML) models face challenges in latency, privacy, and reliability, especially in resource-constrained settings. This paper presents a lightweight anomaly detection framework using TinyML models deployed on microcontroller-based edge devices for real-time physiological signal analysis. DHT11 sensors simulate temperature signals and inference is performed locally using an optimized TensorFlow Lite model. Communication with external systems is handled via Message Queuing Telemetry Transport (MQTT) for efficient data logging and visualization. The framework integrates Tiny Federated Learning (TinyFL) for continual model updates, achieving an average accuracy of 99% while significantly reducing latency and energy consumption across the evaluation rounds. This approach offers a low-latency, scalable and secure solution for decentralized IoMT applications.

To maintain a healthy and balanced lifestyle, it is essential to consume food in proportion to one’s individual needs. The proportions of the food consumed need to be calculated to determine the exact calorie intake or to keep a log. The proposed system, iLog 3.0, automatically determines the volume or quantity of the food item when uploaded using a mobile application, utilizing state-of-the-art object detection and depth estimation techniques for 2D RGB images. The food item will be identified using the Mask R-CNN (Mask Region-Based Convolutional Neural Network) technique. To determine the height of the food item, the MiDaS (Mixed Depth and Scale) technique is employed to generate a depth map, from which the height is subsequently determined. A high success rate has been achieved, and quantification is accurate compared to the previously used models.

CNN models used in disease management systems for disease detection are typically stored and executed on remote cloud servers, communicating with sensors over the internet. This paradigm raises concerns about data security and reliability, especially in remote farmlands. Edge computing introduces a computing layer closer to sensory nodes, enabling more reliable communication and faster results. These edge devices are often resource constrained and cannot run CNN models that need high memory and computation. Since plant disease semantics are simple and consistent across channels, Chroma-Sense proposes processing the R, G, and B channels independently using the same feature extractor. Individual processing reduces the width of the feature extractor, lowering RAM usage and number of computations performed. In addition, reusing the same feature extractor minimizes the parameter count, the Flash memory required to store the model, enabling efficient edge deployment. Proposed Chroma-Sense was trained on a subset of the PlantVillage dataset and achieved a 25% reduction in peak RAM usage and a 60% reduction in flash memory when tested on three different edge devices with varying heap and storage capacities.

In recent years, the rapid growth in urbanization and smart cities has been empowered by efficient and intelligent solutions in areas such as transportation, governance, and smart banking. These applications feature large-scale Internet of Things (IoT) deployment of wirelessly connected smart embedded devices with sensors and actuators. Providing adequate energy to power this large number of IoT devices remains a crucial challenge. In this regard, energy harvesting (EH) emerges as a promising approach that converts ambient energy into usable electrical energy, allowing IoT devices to function autonomously and sustainably, thereby reducing maintenance efforts while enhancing the overall system reliability. Although EH systems offer significant advantages, they are also vulnerable to various threats and attacks that underscore the need to design secure and reliable EH solutions. In this article, we comprehensively review the state-of-the-art EH techniques and associated security aspects. We discuss current research on EM techniques, optimization algorithms, and the challenges involved in energy-efficient routing within IoT. Next, we analyze the existing EH methods in two directions-energy extraction and energy storage. In terms of energy extraction from renewable sources, we review the maximum power point tracking (MPPT) algorithms used in IoT devices. Then, we explore the energy storage capabilities of sensor nodes, which are crucial for consistent operation. Furthermore, we discuss the security and reliability mechanisms within IoT EH frameworks, identifying the threats and countermeasures.We conclude the survey by discussing the future research directions and listing a few open problems.