Power-efficient dynamic routing and robust data transmission in mm-wave healthcare WSN using APADBC and RDHSOA

Optimal Routing (OR) and reliable Data Transmission (DT) are essential for improving the network lifetime in millimeter-wave (mm-wave) Healthcare (HC) Wireless Sensor Networks (WSNs); however, existing frameworks such as LEACH, PEGASIS, and meta-heuristic-based approaches primarily focus on routing efficiency while neglecting the high power consumption of Radio Frequency (RF) front-end circuits, including power amplifiers, oscillators, and beamforming modules, which significantly reduce network longevity. This limitation is further aggravated by the lack of integrated solutions that jointly address RF power optimization and dynamic routing in mm-wave healthcare WSNs, leading to reduced network lifetime and unreliable data transmission under dynamic conditions. To overcome this gap, the proposed framework integrates adaptive RF power optimization and dynamic routing to enhance both energy efficiency and communication reliability. To address this limitation, a power-efficient dynamic routing framework is proposed using Adaptive Power Amplifier Drag Bias Control (APADBC) and the Reinforcement Doll Maker Hurwitz Shekel Optimization Algorithm (RDHSOA). The proposed APADBC dynamically adjusts the bias of RF front-end circuits using a drag equation-based adaptive step size, thereby reducing power consumption while maintaining transmission quality, whereas RDHSOA integrates meta-heuristic optimization with reinforcement learning to enable adaptive and reliable routing path selection under dynamic network conditions. The framework further incorporates cluster head selection, efficient node clustering, and a health risk classification module for analyzing patient vital signs. Experimental results demonstrate that the proposed approach achieves a power gain of 1.758 dB compared to 0.7545 dB achieved by existing methods such as APABC, ADPA, SPAAB, and CCABC, indicating an improvement of over 130%, along with reduced latency and energy consumption of 11 J. Furthermore, the proposed DTRAIBNN achieves a classification accuracy of 99.19%, outperforming conventional models such as DRNN, LSTM, GRU, and CNN. These results confirm the superiority of the proposed framework over existing benchmark approaches in terms of power efficiency, routing reliability, and overall data transmission performance in mm-wave healthcare WSN environments.