Advance Research in Power Electronics and Devices

An Energy-Efficient Approximate Dadda Multiplier using Almost Full Adders, Majority-Logic 4:2 Compressors, and Clock Gating

2026-07-01T11:14:21+00:00

Approximate computing has emerged as a key strategy for energy-efficient very large scale integration design, particularly in signal-processing and multimedia applications where modest reductions in arithmetic accuracy are acceptable in exchange for substantial savings in power and delay. This paper presents a low-power 8×8 approximate Dadda multiplier built from approximate, almost-full adders and majority-logic-based 4:2 compressors. The almost full adder reduces transistor count and switching activity relative to a conventional full adder, while the majority-logic 4:2 compressor simplifies the compressor structure and shortens the critical path during partial-product reduction. A modified Dadda reduction format is introduced to lower the maximum output delay and to decrease the number of MOSFETs required compared with conventional Dadda and Wallace-tree multipliers. To improve energy efficiency further, a clock-gating technique is incorporated to suppress unnecessary switching activity in idle portions of the circuit. The design is implemented in Verilog HDL and verified using the Xilinx Vivado Design Suite. Simulation results confirm correct functionality together with reduced propagation delay, lower hardware complexity, and an improved power-delay product, establishing the architecture as well-suited to error-tolerant DSP, image-processing, and edge-computing applications.

Power-Aware Task Scheduling using a MATLAB/Simulink Simulation Process

2026-06-19T11:37:28+00:00

Battery life is a concern in modern devices. Many devices, like smartwatches, IoT sensors, medical monitors, and industrial controllers, run on batteries. When the battery runs out, the device stops working. This paper examines how task scheduling affects power consumption. Two scheduling methods, Rate Monotonic (RM) and Earliest Deadline First (EDF), are studied. A method called Power-Aware EDF (PA-EDF) is also proposed. PA-EDF reduces processor speed when slack time is available. RM, EDF, and PA-EDF were tested using MATLAB/Simulink. This gave a picture of the system and measurable energy results. The Simulink model has three tasks. It also has a power model and an energy tracker. The results show that PA-EDF saves a lot of energy compared to EDF and more compared to RM. PA-EDF makes sure every task finishes on time. The proposed approach was evaluated through simulations developed in MATLAB/Simulink using a model that included multiple tasks, a processor power model, and an energy measurement mechanism. The performance of the scheduling methods was analyzed in terms of energy consumption and the ability to meet task deadlines. The results indicate that PA-EDF reduces power usage more effectively than both RM and standard EDF while maintaining reliable task execution without missing deadlines. The main contribution of the work is the integration of power-saving mechanisms into an existing scheduling method to improve battery efficiency in real-time systems.

IoT-based Human Wearable Epileptic Seizure Alert System

2026-06-25T11:07:52+00:00

Epilepsy is a chronic neurological disorder characterized by recurrent and unpredictable seizures caused by abnormal brain activity. Sudden seizure episodes may lead to injuries, unconsciousness, breathing irregularities, and other critical conditions if timely medical assistance is not provided. Conventional monitoring techniques such as Electroencephalography (EEG) and Electrocardiography (ECG) are commonly used for seizure analysis; however, these systems are often costly, bulky, and unsuitable for continuous daily monitoring outside clinical environments. This paper presents an IoT-based wearable epileptic seizure alert system designed for real-time monitoring and emergency notification. The proposed prototype integrates a compact microcontroller with motion and physiological sensing modules to monitor abnormal body movements, heart rate variations, and oxygen saturation levels associated with seizure conditions. A multi-parameter monitoring approach is utilized to improve the reliability of seizure detection and minimize false triggering. Upon detection of abnormal conditions, the system activates an emergency alert mechanism that includes an audible alarm, wireless notification to caregivers, and real-time location sharing for rapid assistance. The wearable device is designed to be lightweight, portable, energy efficient, and suitable for continuous health monitoring applications. The proposed system demonstrates the potential of IoT-enabled wearable healthcare technology for improving patient safety, remote monitoring, and emergency response in epilepsy management.