Resource-constrained Devices Research Articles

A novel scheme for enhanced content integrity, authentication, and privacy in information centric network using lightweight blockchain-based homomorphic integrity and authentication

Information-centric networks (ICNs) face challenges in ensuring content integrity and authentication while preserving user privacy. Homomorphic encryption has been widely used to protect content privacy within ICNs. Blockchain technology is widely applied in ICNs to ensure content integrity. Conventional blockchain-based authentication schemes rely on computationally expensive homomorphic encryption to perform secure data transmission. Also, the existing cryptography methods comprise confidentiality, and integrity while sharing content in ICNs. This paper proposes a novel lightweight blockchain-based homomorphic integrity and authentication (Light-BHIA) scheme to increase the integrity and privacy of content within ICNs. This scheme leverages the transparency and immutability of blockchain, the privacy-preserving properties of homomorphic encryption, and decentralized trust mechanisms to achieve secure and verifiable content delivery. The proposed work utilizes a lightweight homomorphic encryption scheme specifically designed for ICNs, reducing computational burdens by using resource-constrained devices. Light-BHIA uses the permissioned blockchain with an efficient consensus mechanism for increasing the trustworthiness of content delivery within ICNs. Authorized recipients can perform integrity and authentication checks on the encrypted content without decryption, maintaining confidentiality. The comparative study reveals proposed scheme achieved a 25% reduction in registration delay and confirms higher integrity, privacy, and confidentiality.

Efficient vehicular counting via privacy-aware aggregation network

Abstract Vehicle counting is crucial for effective road planning and traffic management. Despite significant advancements that have been achieved with the development of deep learning technology, current counting models rely on large-scale parameters and substantial computational resources, which limits their practical application. Additionally, these methods are typically trained on large centralized datasets, which may result in inefficiencies for resource-constrained devices. Furthermore, inadequate privacy protection poses potential risks of personal information leakage. To address these issues, we introduce a lightweight counting network, privacy-aware aggregation network (PANet) for real-world application in this paper. In PANet, a pyramid feature enhancement module is built to aggregate multi-scale information and enhance key representation, while also optimizing the channel-wise output of the model to reduce computational complexity. Furthermore, a federated learning framework is implemented to distribute the computational load and safeguard user privacy. Experimental results on a wide range of counting benchmarks demonstrate the superior efficiency and accuracy of PANet. The code is available at https://github.com/sdut-jacheng/PANet.

SIMECK-T: An Ultra-Lightweight Encryption Scheme for Resource-Constrained Devices

The Internet of Things produces vast amounts of data that require specialized algorithms in order to secure them. Lightweight cryptography requires ciphers designed to work on resource-constrained devices like sensors and smart things. A new encryption scheme is introduced based on a blend of the best-performing algorithms, SIMECK and TEA. A selection of software-oriented Addition–Rotation–XOR (ARX) block ciphers are augmented with a dynamic substitution security layer. The performance is compared against other lightweight approaches. The US National Institute of Standards and Technology (NIST) SP800-22 Statistical Test Suite for Random and Pseudorandom Number Generators for Cryptographic Applications and the German AIS.31 of the Federal Office for Information Security (BSI) are used to validate the output of the proposed encryption scheme. The law of iterated logarithm (LIL) for randomness is verified in all three forms. The total variance (TV), the Hellinger Distance (HD), and the root-mean-square deviation (RMSD) show values smaller than the required limit for 10.000 sequences of ciphertext. The performance evaluation is analyzed on a Raspberry PICO 2040. Several security metrics are compared against other ciphers, like χ2 and encryption quality (EQ). The results show that SIMECK-T is a powerful and fast, software-oriented, lightweight cryptography solution.

A Review on Secure Authentication Mechanisms for Mobile Security

Cybersecurity, complimenting authentication, has become the backbone of the Internet of Things. In the authentication process, the word authentication is of the utmost importance, as it is the door through which both Mr. Right Guy and Mr. Wrong Guy can pass. It is the key to opening the most important and secure accounts worldwide. When authentication is complete, surely there will be passwords. Passwords are a brain-confusing option for the user to choose when making an account during the registration/sign-up process. Providing reliable, effective, and privacy-preserving authentication for individuals in mobile networks is challenging due to user mobility, many attack vectors, and resource-constrained devices. This review paper explores the transformation and modern mobile authentication schemes, categorizing them into password, graphical, behavioral, keystroke, biometric, touchscreen, color, and gaze-based methodologies. It aims to examine the strengths and limitations focused on challenges like security and usability. Standard datasets and performance evaluation measures are also discussed. Finally, research gaps and future directions in this essential and emerging area of research are discussed.

Deep-Learning Framework for Efficient Real-Time Speech Enhancement and Dereverberation

Deep learning has revolutionized speech enhancement, enabling impressive high-quality noise reduction and dereverberation. However, state-of-the-art methods often demand substantial computational resources, hindering their deployment on edge devices and in real-time applications. Computationally efficient approaches like deep filtering and Deep Filter Net offer an attractive alternative by predicting linear filters instead of directly estimating the clean speech. While Deep Filter Net excels in noise reduction, its dereverberation performance remains limited. In this paper, we present a generalized framework for computationally efficient speech enhancement and, based on this framework, identify an inherent constraint within Deep Filter Net that hinders its dereverberation capabilities. We propose an extension to the Deep Filter Net framework designed to overcome this limitation, demonstrating significant improvements in dereverberation performance while maintaining competitive noise-reduction quality. Our experimental results highlight the potential of this enhanced framework for real-time speech enhancement on resource-constrained devices.

A Low-Power General Matrix Multiplication Accelerator with Sparse Weight-and-Output Stationary Dataflow.

General matrix multiplication (GEMM) in machine learning involves massive computation and data movement, which restricts its deployment on resource-constrained devices. Although data reuse can reduce data movement during GEMM processing, current approaches fail to fully exploit its potential. This work introduces a sparse GEMM accelerator with a weight-and-output stationary (WOS) dataflow and a distributed buffer architecture. It processes GEMM in a compressed format and eliminates on-chip transfers of both weights and partial sums. Furthermore, to map the compressed GEMM of various sizes onto the accelerator, an adaptable mapping scheme is designed. However, the irregular sparsity of weight matrices makes it difficult to store them in local buffers with the compressed format; denser vectors can exceed the buffer capacity, while sparser vectors may lead to the underutilization of buffers. To address this complication, this work also proposes an offline sparsity-aware shuffle strategy for weights, which balances the utilization of distributed buffers and minimizes buffer waste. Finally, a low-cost sparse computing method is applied to the WOS dataflow with globally shared inputs to achieve high computing throughput. Experiments with an FPGA show that the proposed accelerator achieves 1.73× better computing efficiency and 1.36× higher energy efficiency than existing approaches.

Energy-Efficient Distributed Edge Computing to Assist Dense Internet of Things

The Internet of Things (IoT) represents a rapidly growing field, where billions of intelligent devices are interconnected through the Internet, enabling the seamless sharing of data and resources. These smart devices are typically employed to sense various environmental characteristics, including temperature, motion of objects, and occupancy, and transfer their values to the nearest access points for further analysis. The exponential growth in sensor availability and deployment, powered by recent advances in sensor fabrication, has greatly increased the complexity of IoT network architecture. As the market for these sensors grows, so does the problem of ensuring that IoT networks meet high requirements for network availability, dependability, flexibility, and scalability. Unlike traditional networks, IoT systems must be able to handle massive amounts of data generated by various and frequently-used resource-constrained devices, while ensuring efficient and dependable communication. This puts high constraints on the design of IoT, mainly in terms of the required network availability, reliability, flexibility, and scalability. To this end, this work considers deploying a recent technology of distributed edge computing to enable IoT applications over dense networks with the announced requirements. The proposed network depends on distributed edge computing at two levels: multiple access edge computing and fog computing. The proposed structure increases network scalability, availability, reliability, and scalability. The network model and the energy model of the distributed nodes are introduced. An energy-offloading method is considered to manage IoT data over the network energy, efficiently. The developed network was evaluated using a developed IoT testbed. Heterogeneous evaluation scenarios and metrics were considered. The proposed model achieved a higher energy efficiency by 19%, resource utilization by 54%, latency efficiency by 86%, and reduced network congestion by 92% compared to traditional IoT networks.

Cryptography in IoT: Securing the Next Generation of Connected Devices

The Internet of Things (IoT) has revolutionized connectivity across diverse industries, linking billions of devices for optimized operations, data-driven insights, and enhanced consumer experiences. However, IoT's vast scale, heterogeneity, and resource-constrained devices introduce severe security challenges, with data integrity and privacy at risk. This white paper explores the application of cryptographic methods in securing IoT, addressing the unique constraints IoT presents, and highlighting recent advancements in lightweight cryptography, key management, and authentication protocols tailored to IoT requirements. Key words: IoT Security, Cryptography, Lightweight Encryption, Key Management, Authentication, Secure Communication

TinyML and edge intelligence applications in cardiovascular disease: A survey.

Tiny machine learning (TinyML) and edge intelligence have emerged as pivotal paradigms for enabling machine learning on resource-constrained devices situated at the extreme edge of networks. In this paper, we explore the transformative potential of TinyML in facilitating pervasive, low-power cardiovascular monitoring and real-time analytics for patients with cardiac anomalies, leveraging wearable devices as the primary interface. To begin with, we provide an overview of TinyML software and hardware enablers, accompanied by an examination of networking solutions such as Low-power Wide area network (LPWAN) that facilitate the seamless deployment of TinyML frameworks. Following this, we delve into the methodologies of knowledge distillation, quantization, and pruning, which represent the cornerstone strategies for optimizing machine learning models to operate efficiently within resource-constrained environments. Furthermore, our discussion extends to the role of efficient deep neural networks tailored specifically for cardiovascular monitoring on wearable devices with limited computational resources. Through a comprehensive review, we analyze the applications of prominent artificial neural network architectures including Convolutional Neural Networks (CNNs), Autoencoders, Deep Belief Networks (DBNs), and Transformers in the domain of Electrocardiogram (ECG) analytics, shedding light on their efficacy and potential in advancing healthcare technology.

Efficient Multi-Task Training with Adaptive Feature Alignment for Universal Image Segmentation.

Universal image segmentation aims to handle all segmentation tasks within a single model architecture and ideally requires only one training phase. To achieve task-conditioned joint training, a task token needs to be used in the multi-task training to condition the model for specific tasks. Existing approaches generate the task token from a text input (e.g., "the task is panoptic"). However, such text-based inputs merely serve as labels and fail to capture the inherent differences between tasks, potentially misleading the model. In addition, the discrepancy between visual and textual modalities limits the performance gains in existing text-involved segmentation models. Nevertheless, prevailing modality-alignment methods rely on large-scale uni-modal encoders for both modalities and an extremely large amount of paired data for training, and therefore it is hard to apply these existing models to lightweight segmentation models and resource-constrained devices. In this paper, we propose Adaptive Feature Alignment (AFA) integrated with a learnable task token to address these issues. The learnable task token automatically captures inter-task differences from both image features and text queries during training, providing a more effective and efficient solution than a predefined text-based token. To efficiently align the two modalities without introducing extra complexity, we reconsider the differences between a text token and an image token and replace image features with class-specific means in our proposed AFA. We evaluate our model performance on the ADE20K and Cityscapes datasets. Experimental results demonstrate that our model surpasses baseline models in both efficiency and effectiveness, achieving state-of-the-art performance among segmentation models with a comparable amount of parameters.

Enhancing cognitive radio networks for education systems: A machine learning approach to optimized spectrum sensing in remote learning environments

The present study focuses on improving Cognitive Radio Networks (CRNs) based on applying machine learning to spectrum sensing in remote learning scenarios. Remote education requires connection dependability and continuity that can be affected by the scarcity of the amount of usable spectrum and suboptimal spectrum usage. The solution for the proposed problem utilizes deep learning approaches, namely CNN and LSTM networks, to enhance the spectrum detection probability (92% detection accuracy) and consequently reduce the number of false alarms (5% false alarm rate) to maximize spectrum utilization efficiency. By developing the cooperative spectrum sensing where many users share their data, the system makes detection more reliable and energy-saving (achieving 92% energy efficiency) which is crucial for sustaining stable connections in educational scenarios. This approach addresses critical challenges in remote education by ensuring scalability across diverse network conditions and maintaining performance on resource-constrained devices like tablets and IoT sensors. Combining CRNs with new technologies like IoT and 5G improves their capabilities and allows these networks to meet the constantly changing loads of distant educational systems. This approach presents another prospect to spectrum management dilemmas in that education delivery needs are met optimally from any STI irrespective of the availability of resources in the locale. The results show that together with machine learning, CRNs can be considered a viable path to improving the networks’ performance in the context of remote learning and advancing the future of education in the digital environment. This work also focuses on how machine learning has enabled the enhancement of CRNs for education and provides robust solutions that can meet the increasing needs of online learning.

An AIoT System for Earthquake Early Warning on Resource Constrained Devices

An AIoT System for Earthquake Early Warning on Resource Constrained Devices

Enhancing deep convolutional neural network models for orange quality classification using MobileNetV2 and data augmentation techniques

This study introduces significant improvements in the construction of deep convolutional neural network models for classifying agricultural products, specifically oranges, based on their shape, size, and color. Utilizing the MobileNetV2 architecture, this research leverages its efficiency and lightweight nature, making it suitable for mobile and embedded applications. Key techniques such as depthwise separable convolutions, linear bottlenecks, and inverted residuals help reduce the number of parameters and computational load while maintaining high performance in feature extraction. Additionally, the study employs comprehensive data augmentation methods, including horizontal and vertical flips, grayscale transformations, hue adjustments, brightness adjustments, and noise addition to enhance the model's robustness and generalization capabilities. The proposed model demonstrates superior performance, achieving an overall accuracy of 99.53%∼100% with nearly perfect precision, recall of 95.7%, and F1-score of 94.6% for both “orange_good” and “orange_bad” classes, significantly outperforming previous models which typically achieved accuracies between 70% and 90%. While the classification performance was near-perfect in some aspects, there were minor errors in specific detection tasks. The confusion matrix shows that the model has high sensitivity and specificity, with very few misclassifications. Finally, this study highlights the practical applicability of the proposed model, particularly its easy deployment on resource-constrained devices and its effectiveness in agricultural product quality control processes. These findings affirm the model in this research as a reliable and highly efficient tool for agricultural product classification, surpassing the capabilities of traditional models in this field.

Intelligent Remote Sensing Scene Classification Model for On-Board Training of Resource-Constrained Devices

Remote Sensing Scene Classification (RSSC) is the distinctive classification of remote sensing images into numerous classes of scene classifications based on the image content. RSSC plays a significant role in several domains, like land mapping, agriculture, and the classification of disaster-prone regions. The Internet of Things (IoT) is a dynamic global network of devices, for example, vehicles, sensors, actuators, surveillance cameras, etc. These interconnected objects were distinctively recognizable and they could separately transfer and obtain valuable data through the network. However, satellite images were frequently degraded and blurred owing to aerosol dispersion under haze, fog, and other weather circumstances, decreasing the color fidelity and contrast of the image. To use effectual RSSC in real-time, widespread researchers concentrate on creating aerospace image processing systems, like airborne or spaceborne systems. Recently, with the quick improvement of deep learning (DL) and Machine learning (ML) techniques, the performance of RSSC has significantly developed owing to the hierarchical feature representation learning. Both technique has greater achievement in the domain of image scene classification. This study presents a Leveraging Tiny Convolutional Neural Networks with a Water Cycle Algorithm for Remote Sensing Scene Classification (LTCNN-WCRSSC) model. The LTCNN-WCRSSC technique is designed for efficient RSS classification in resource-constrained devices with on-board training capabilities. At first, the LTCNN-WCRSSC model applies image processing using a median filter (MF) to eliminate the noise. Next, the feature extraction process can be exploited by the ConvNeXt-Tiny method. For the RSSC model, the spatiotemporal attention bidirectional long short-term memory (STA-BiLSTM) technique is performed. Eventually, the water cycle algorithm (WCA)-based hyperparameter choice process can be performed to optimize the classification results of the STA-BiLSTM algorithm. The experimental evaluation of the LTCNN-WCRSSC technique takes place using a benchmark image dataset. The stimulated results indicated the superior performances of the LTCNN-WCRSSC model over other approaches.

Advanced Pedestrian Distance Estimation for ADAS with Canny Edge Detection and Stereo Vision

Pedestrian detection is a vital aspect of Advanced Driver Assistance Systems (ADAS), crucial for ensuring driving safety and minimizing collision risks. While detecting pedestrians is important, it must be paired with precise distance estimation to create a robust safety solution. Stereovision cameras are well-regarded for their effectiveness and affordability in measuring depth through disparity between two images. Despite this, research on pedestrian distance estimation using only stereovision remains sparse, with many studies relying on computationally heavy dense depth maps. This paper proposes an innovative method for computing object-level disparity specifically for pedestrian detection using stereo cameras. The approach integrates Canny edge detection with ORB (Oriented FAST and Rotated BRIEF) feature matching to efficiently identify and track keypoints within pedestrian bounding boxes. This method not only improves the accuracy of distance estimation but also reduces computational demands, making it suitable for real-time applications. The approach was thoroughly tested on a Raspberry Pi 4, a resource-constrained device, and achieved promising results, demonstrating its potential for practical use in ADAS.

Knowledge distillation approach for skin cancer classification on lightweight deep learning model.

Over the past decade, there has been a global increase in the incidence of skin cancers. Skin cancer has serious consequences if left untreated, potentially leading to more advanced cancer stages. In recent years, deep learning based convolutional neural network have emerged as powerful tools for skin cancer detection. Generally, deep learning approaches are computationally expensive and require large storage space. Therefore, deploying such a large complex model on resource-constrained devices is challenging. An ultra-light and accurate deep learning model is highly desirable for better inference time and memory in low-power-consuming devices. Knowledge distillation is an approach for transferring knowledge from a large network to a small network. This small network is easily compatible with resource-constrained embedded devices while maintaining accuracy. The main aim of this study is to develop a deep learning-based lightweight network based on knowledge distillation that identifies the presence of skin cancer. Here, different training strategies are implemented for the modified benchmark (Phase 1) and custom-made model (Phase 2) and demonstrated various distillation configurations on two datasets: HAM10000 and ISIC2019. In Phase 1, the student model using knowledge distillation achieved accuracies ranging from 88.69% to 93.24% for HAM10000 and from 82.14% to 84.13% on ISIC2019. In Phase 2, the accuracies ranged from 88.63% to 88.89% on HAM10000 and from 81.39% to 83.42% on ISIC2019. These results highlight the effectiveness of knowledge distillation in improving the classification performance across diverse datasets and enabling the student model to approach the performance of the teacher model. In addition, the distilled student model can be easily deployed on resource-constrained devices for automated skin cancer detection due to its lower computational complexity.

Approximate CNN Hardware Accelerators for Resource Constrained Devices

Approximate CNN Hardware Accelerators for Resource Constrained Devices

A multi-agent reinforcement learning based approach for automatic filter pruning

Deep Convolutional Neural Networks (DCNNs), due to their high computational and memory requirements, face significant challenges in deployment on resource-constrained devices. Network Pruning, an essential model compression technique, contributes to enabling the efficient deployment of DCNNs on such devices. Compared to traditional rule-based pruning methods, Reinforcement Learning(RL)—based automatic pruning often yields more effective pruning strategies through its ability to learn and adapt. However, the current research only set a single agent to explore the optimal pruning rate for all convolutional layers, ignoring the interactions and effects among multiple layers. To address this challenge, this paper proposes an automatic Filter Pruning method with a multi-agent reinforcement learning algorithm QMIX, named QMIX_FP. The multi-layer structure of DCNNs is modeled as a multi-agent system, which considers the varying sensitivity of each convolutional layer to the entire DCNN and the interactions among them. We employ the multi-agent reinforcement learning algorithm QMIX, where individual agent contributes to the system monotonically, to explore the optimal iterative pruning strategy for each convolutional layer. Furthermore, fine-tuning the pruned network using knowledge distillation accelerates model performance improvement. The efficiency of this method is demonstrated on two benchmark DCNNs, including VGG-16 and AlexNet, over CIFAR-10 and CIFAR-100 datasets. Extensive experiments under different scenarios show that QMIX_FP not only reduces the computational and memory requirements of the networks but also maintains their accuracy, making it a significant advancement in the field of model compression and efficient deployment of deep learning models on resource-constrained devices.

IoT Security: Botnet Detection Using Self-Organizing Feature Map and Machine Learning

The rapid advancement of Internet of Things (IoT) technology has created potential for progress in various aspects of life. However, the increasing number of IoT devices also raises the risk of cyberattacks, particularly IoT botnets often exploited by attackers. This is largely due to the limitations of IoT devices, such as constraints in capacity, power, and memory, necessitating an efficient detection system. This study aims to develop a resource-efficient botnet detection system by using the Self-Organizing Feature Map (SOFM) dimensionality reduction method in combination with machine learning algorithms. The proposed method includes a feature engineering process using SOFM to address high-dimensional data, followed by classification with various machine learning algorithms. The experiments evaluate performance based on accuracy, sensitivity, specificity, False Positive Rate (FPR), and False Negative Rate (FNR). Results show that the Decision Tree algorithm achieved the highest accuracy rate of 97.24%, with a sensitivity of 0.9523, specificity of 0.9932, and a fast execution time of 100.66 seconds. The use of SOFM successfully reduced memory consumption from 3.08 GB to 923MB. Experimental results indicate that this approach is effective for enhancing IoT security in resource-constrained devices.

Federated Collaborative Learning with Sparse Gradients for Heterogeneous Data on Resource-Constrained Devices.

Federated learning enables devices to train models collaboratively while protecting data privacy. However, the computing power, memory, and communication capabilities of IoT devices are limited, making it difficult to train large-scale models on these devices. To train large models on resource-constrained devices, federated split learning allows for parallel training of multiple devices by dividing the model into different devices. However, under this framework, the client is heavily dependent on the server's computing resources, and a large number of model parameters must be transmitted during communication, which leads to low training efficiency. In addition, due to the heterogeneous distribution among clients, it is difficult for the trained global model to apply to all clients. To address these challenges, this paper designs a sparse gradient collaborative federated learning model for heterogeneous data on resource-constrained devices. First, the sparse gradient strategy is designed by introducing the position Mask to reduce the traffic. To minimize accuracy loss, the dequantization strategy is applied to restore the original dense gradient tensor. Second, the influence of each client on the global model is measured by Euclidean distance, and based on this, the aggregation weight is assigned to each client, and an adaptive weight strategy is developed. Finally, the sparse gradient quantization method is combined with an adaptive weighting strategy, and a collaborative federated learning algorithm is designed for heterogeneous data distribution. Extensive experiments demonstrate that the proposed algorithm achieves high classification efficiency, effectively addressing the challenges posed by data heterogeneity.