Steganography is considered the first line of defense in information security as it hides a secret message (payload) inside an innocent looking file (container) to transfer the payload under the adversary's nose without noticing it. Steganographic systems only use the container to hide the payload. In this paper, we present a steganographic system that uses the container not only to hide the payload, but also to give misleading information to the adversary. To achieve this goal, we use quick response (QR) code as a container. QR codes generated by our proposed system can carry its ordinary message in addition to the payload. Anyone can read the message, but the payload can only be obtained using a secret key. The message and the payload are unrelated; i.e. any message can be generated regardless of the payload and vise versa. We can take advantage of that by generating a message that gives misleading information to the adversary. We test the proposed system and show that the generated QR code is (valid) i.e indistinguishable from an ordinary QR code which makes it look innocent and less susceptible to an adversary's attack. Moreover, it is space-efficient, has an acceptable level of noise immunity and is prone to steganalysis attacks.
1 April 2024 in Journal of King Saud University Computer and Information Sciences
In recent years, facial recognition technology has become increasingly integrated into society, making privacy protection crucial. Previous techniques offered minimal secrecy safeguards through simple obscuration methods. This paper addresses the strict privacy requirements of face image data by developing a novel framework that synergistically integrates Generative Adversarial Networks (GANs), clustering algorithms, and Blockchain technology. The methodology proposes a cutting-edge Privacy-Preserving Self-Attention GAN (PPSA-GAN) to generate realistic synthetic facial imagery. An integrated mini-batch K-means clustering algorithm anonymizes these images into distinct groupings, maximizing privacy preservation. Blockchain integration complements the system by fortifying trust through decentralized ledgers for transparent yet secure data storage and auditing. Rigorous benchmarking on the CelebA dataset confirms the PPSA-GAN architecture’s state-of-the-art performance, attaining an impressive Inception Score of 13.99 and a Fréchet Inception Distance of 35.50. The mini-batch clustering forms 125 distinct clusters, effectively anonymizing facial attributes within the synthetic images. Blockchain integration further bolsters privacy assurances via tamper-proof historical records, showcasing precision, recall, F1-score, and accuracy values of 0.948, 0.938, 0.943, and 0.947, respectively. This multifunctional framework represents a novel contribution, fostering an ethical technological ecosystem that balances progress and privacy. Prospective deployment horizons encompass identity verification, surveillance infrastructure, and augmentation of medical image repositories, seeding an enlightening future for facial recognition domains.
According to the World Health Organization (WHO), congenital heart disorders (CHDs) impact a significant proportion of babies worldwide, with prevalence rates ranging from 0.8% to 1.2%. Phonocardiography (PCG) is the leading non-invasive method for detecting congenital heart defects (CHDs), offering important information about cardiac signals i.e. S1, S2, S3, and S4 and heartbeat patterns. This research study focuses on developing a strong binary classification system for congenital heart diseases (CHDs) utilizing deep neural networks. The system will be trained on a combined dataset that includes both local and publicly available repositories. The local dataset (LD) consists of 583 signals containing both normal and aberrant PCG recordings, while the public dataset (PD) obtained from Michigan University consists of 23 PCG recordings. In order to achieve consistency and compatibility, both datasets are subjected to down-sampling, resulting in a frequency of 8 kHz. A well-engineered band-pass filter efficiently removes signals outside the 20-650 Hz range, allowing for exclusive processing of the desired frequencies. The signals are divided into segments, each lasting exactly 4 seconds. The study utilizes data augmentation techniques, specifically pitch-shifting, to boost the model’s robustness. This is accomplished by implementing a 1D convolutional neural network (CNN). The most notable outcomes are achieved in case C, exhibiting a sensitivity of 99.0%, specificity of 98.0%, F1 score of 98.56%, precision of 98.57%, and an accuracy of 98.56%. This study enhances the progress of automated techniques for detecting congenital heart disease (CHD), demonstrating the potential of deep neural networks in precision medicine for pediatric cardiology.
Threat detection in a Cyber-Physical System (CPS) platform is a key feature of ensuring the reliability and security of these connected methods, but digital elements interface with the physical world. CPS platforms are popular in sectors like healthcare, industrial automation, smart cities, and transportation making them vulnerable to different cyber-attacks. Threat detection in CPS contains the detection and mitigation of cybersecurity risks, which disrupt physical processes, compromise data integrity, and potentially cause safety concerns. Machine learning (ML) and deep learning (DL) systems are exploited for detecting anomalies by learning the normal behaviour forms of the CPS and recognizing deviations. This study presents an Automated Threat Detection using the Flamingo Search Algorithm with Optimal Deep Learning (ATD-FSAODL) technique in a CPS environment. Initially, the ATD-FSAODL technique applies FSA-based feature subset selection to elect the better group of features. In addition, the ATD-FSAODL technique makes use of a modified Elman Spike Neural Network (MESNN) model for threat recognition and classification. Finally, the slime mold algorithm (SMA) is used for the optimal selection of the parameters related to the MESNN approach to ensure that the threat detection rate is improved. To estimate the solution of the ATD-FSAODL technique, a sequence of simulations can be carried out on benchmark databases. The performance values portray the capable solution of the ATD-FSAODL methodology with other methods with a maximum accuracy of 99.58%, precision of 99.58%, recall of 99.58%, F-score of 99.58%, and MCC of 99.16%.
The exponential increase in the Internet of Things (IoT) has affected the cloud computing with increase transmission latency and network overhead for real-time applications. Cloud-fog computing paradigm tackle these limitations by moving computational services closer to the network edge i.e., fog nodes, enhancing the speed of real-time applications. This architecture, with its dynamic computing environment and diverse IoT devices and tasks, demands a reliable and energy-efficient communication network. Joint optimization of computation offloading and task scheduling is a primary challenge, as it involves offloading tasks to optimal computational resources and scheduling them in an efficient order for operational efficacy. While offloading tasks to fog nodes reduces delay but raises energy utilization, offloading them to cloud servers reduces energy usage but raises computational costs and latency. Additionally, inefficient order of task execution (executing lower priority jobs before higher priority tasks) can disrupt system stability and reliability. Therefore, an effective joint optimal computation offloading and task scheduling strategy is essential. To this end, we propose a Multi-objective Arithmetic Optimization-based joint computation offloading and task scheduling algorithm, aiming to minimize energy consumption and transmission latency. Extensive simulations in MATLAB demonstrate the efficacy of the proposed algorithm in terms of designated optimization objectives.