Segmented Regressive Stochastic Deep Perceptive Network for Performance Analytics on Industrial IoT Protocol in Data Transmission
International journal of intelligent engineering and systems · 2025
The Internet of Things (IoT) paradigm interconnects the Industrial application has the ability to compile, transfer and interpret data in efficient way.The major factors in this transformation are the adoption of lightweight, low-power protocols that facilitate seamless communication over wide area coverage.Edge computing involves data processing closer than a centralized centre for data to enhance the effectiveness of IoT systems.With the incorporation of Industrial Internet of Things technologies (IIoT), the demand for effective data communication becomes a significant challenge due to the energy and the limited communication bandwidth.Traditional methods are not wellsuited for improving the performance of IIoT protocol in industrial environments.In this paper, a Segmented Regressive Stochastic Gradient Deep Multilayer Perceptron Neural Network (SRSGDMPNN) is introduced to enhance the performance of IIoT Protocols during data communication.Deep Multilayer Perceptron Neural Network is composed of multiple layers, including one input layer, two hidden layers, and one output layer.Thus the result of SRSGDMPNN shows higher classification accuracy by 96% for 500 IoT devices and 172 (packets/sec) throughput for 10000 data packets than conventional methods.The energy consumption is minimized by 12 joules for 500 IoT devices.Also, the latency and jitter are reduced by 38.9 ms and 21 ms concerning the 10000 data packets when compared to the state-of-art methods.