Journal of Advancement in Electronics Signal Processing

Joint Encryption and Reversible Watermarking Scheme for Secure Data Authentication and Integrity Verification

Madhuri Mohanrao Karad — Sat, 28 Mar 2026 00:00:00 +0000

Reversible watermarking, distinguished from conventional watermarking by its capacity to restore the original host content without residual distortion upon authenticated watermark extraction, has emerged as a critical technology for applications demanding lossless data fidelity alongside provenance authentication—including medical imaging, forensic evidence management, satellite remote sensing, and legal document processing. This paper presents a novel joint encryption and reversible watermarking scheme that integrates homomorphic encryption properties with reversible data hiding based on difference expansion (DE) and histogram shifting (HS) techniques to provide simultaneous data confidentiality, authentication, and integrity verification without sacrificing content fidelity upon authorised extraction. The proposed scheme operates in the encrypted domain, enabling cloud-based service providers to embed authentication watermarks into encrypted content without requiring decryption access, preserving content confidentiality from intermediate processing entities. The joint optimisation framework minimises the total distortion introduced during embedding under constraints of watermark capacity and computational efficiency. Authentication is implemented through a hierarchical scheme combining SHA-3 content hashing with an embedded authentication code, enabling both global content authentication and localised tamper detection with pixel-level granularity. Experimental evaluation on medical imaging datasets, including the CheXpert chest radiograph database and ISIC skin lesion database, demonstrates lossless content recovery with embedding capacities exceeding 1.2 bits per pixel at PSNR values above 50 dB during transmission. The scheme achieves false acceptance rates below 10^-6 and false rejection rates below 10^-4 for tamper detection, substantially improving upon state-of-the-art reversible authentication watermarking approaches.

Design for Multi-vital Health Parameters by Non-invasive Method

Naik D. C., Rudresh H. Karjigi, Kavana M., Mounika M. S., Rishitha C. E. — Sat, 31 Jan 2026 00:00:00 +0000

The growing need for continuous, non-invasive health monitoring has driven rapid advancements in technologies capable of tracking multiple vital physiological parameters in real-time. This project introduces a non-invasive multi-vital health parameter monitoring system that measures glucose levels, hemoglobin concentration, pulse rate, and body temperature without the need for invasive blood sampling. The system is built around two core sensing components: the MAX30102 optical sensor and the DS18B20 digital temperature sensor. The MAX30102 sensor utilizes multi-wavelength Photoplethysmography (PPG) to estimate glucose levels, analyze hemoglobin concentration, and monitor pulse rate. It operates by detecting variations in light absorption caused by blood flow and tissue composition, enabling accurate physiological assessments from the skin’s surface. Complementing this, the DS18B20 sensor provides precise and stable body temperature readings through direct contact with the skin, ensuring reliable thermal data. Together, these sensors feed data into a processing unit that applies optical absorption principles and Digital Signal Processing (DSP) algorithms to extract meaningful insights. This approach eliminates the discomfort and risks associated with traditional invasive methods, offering a safer and more user-friendly alternative for health tracking. The system is designed for real-time monitoring, making it ideal for chronic disease management, wellness applications, and preventive healthcare.

Design and Performance Analysis of a Low-Power Multistage Amplifier for Portable Applications using LTspice

Md. Ali, ASM Shamim Hasan, Md. Sohel Rana, Md. Sumon Ali, Syed Tohabbul Murshed — Tue, 31 Mar 2026 00:00:00 +0000

This study presents the design and simulation of a low-power multistage amplifier intended for modern portable electronic systems. With the rapid expansion of wearable sensors, biomedical monitoring devices, and Internet-of-Things (IoT) technologies, there is a growing demand for analogue circuits capable of delivering high performance while maintaining minimal power consumption. Many of these devices operate using small batteries or energy-harvesting sources, making energy efficiency a critical design requirement. In such systems, analogue amplifiers serve as essential components because they amplify weak electrical signals generated by sensors so that the signals can be accurately processed by subsequent electronic stages. Conventional amplifier designs generally emphasise high gain and wide bandwidth; however, these improvements often lead to higher power dissipation, which is undesirable for portable applications. Consequently, recent research has focused on developing amplifier architectures that preserve strong amplification performance while reducing energy usage. Low-power amplifier design has therefore become a significant area of interest in analogue integrated circuit research. The proposed amplifier architecture consists of three functional stages. The first stage is a differential input stage that performs the initial signal amplification while improving noise rejection and signal accuracy. The second stage operates as the main gain stage, providing substantial voltage amplification while maintaining stability and low power consumption. The third stage functions as an output buffer that isolates the amplifier from the load and enhances its ability to drive external circuits without degrading performance. Simulations performed in the LTspice environment with a 1.8 V supply demonstrate that the amplifier achieves approximately 95 dB voltage gain, a gain-bandwidth product near 1.1 MHz, and total power consumption below 200 µW, indicating a well-balanced and energy-efficient design.

A Chameleon-inspired Biomimetic Sensing and Control Framework for Mood Change Detection and Regulation

Suriyaprabha Ka, Vikramarajan S. S. — Wed, 25 Feb 2026 00:00:00 +0000

The Autonomic Nervous System (ANS) is the major system governing human emotional states and moods, which can be measured using physiological parameters such as Heart Rate Variability (HRV), Galvanic Skin Response (GSR), skin temperature, and breathing patterns. Studies have confirmed that multimodal physiological sensing is a reliable, non-invasive, and real-time approach for monitoring emotions and stress, further enhanced by advanced signal processing and machine learning algorithms. Breathing is a critical component of these parameters, which affect autonomic balance and serve as a major regulator in feedback-controlled emotional processes. Chameleons are exemplary of rapid and dynamic color changes caused by emotional arousal, external forces, and social stimuli through interlinked neural-hormonal mechanisms. Based on this natural resilience, this work proposes a biomimetic chameleon model for human mood detection and control. This approach combines various physiological sensors with a control mechanism that adjusts the intensity of intervention based on the mood changes. Unlike conventional recognition-oriented classifiers, the model emphasizes closed-loop control through subtle intervention tools such as paced breathing exercises, chromatic visual stimuli, audio cues, and yoga responses. Experimental findings indicate improved mood stability, reduced stress indicators, and enhanced affective flexibility, making it a promising approach for active mental health and wellness applications.

A Review on Artificial Intelligence Applications in Modern Signal Processing Systems

Vidhika V. Kagale, Pooja A. Magdum, Madhuri R. Jadhav — Thu, 29 Jan 2026 00:00:00 +0000

Signal processing is a core discipline in electronics and communication engineering that deals with the acquisition, analysis, transformation, and interpretation of signals such as speech, images, biomedical signals, and sensor data. Conventional signal processing techniques, including Fourier analysis, filtering, and linear system modelling, have been widely adopted due to their solid mathematical foundations and computational efficiency. While these methods perform well for structured and stationary signals, they often struggle with real-world data that is noisy, non-linear, time-varying, and high-dimensional. The rapid growth in data complexity and volume has driven the need for more adaptive and intelligent signal processing approaches. Artificial Intelligence (AI), particularly machine learning and deep learning, has emerged as an effective solution to address the limitations of traditional methods. Machine learning algorithms enable systems to automatically learn patterns from data, enhancing performance in tasks such as classification, detection, and prediction. Deep learning models, including convolutional and recurrent neural networks, further improve signal processing by automatically extracting meaningful spatial and temporal features from raw signals, reducing dependency on handcrafted features. In addition, advanced signal representation techniques such as wavelet transforms provide joint time-frequency analysis, making them suitable for non-stationary signal processing. The integration of wavelet-based methods with AI models enhances robustness and accuracy in complex and noisy environments. This paper presents a comprehensive review of AI applications in modern signal processing, covering speech and audio processing, image and video analysis, biomedical signal interpretation, smart sensor networks, and predictive maintenance. Key challenges related to computational complexity, data dependency, interpretability, and ethical issues are also discussed. The study concludes that AI-integrated signal processing systems offer intelligent, scalable, and robust solutions for complex real-world engineering applications.