Edge-hosted digital twin with time-series predictive routing for industrial IoT networks

Mass deployments of Wireless Sensor Networks (WSNs) empowered by the Internet of Things (IoT) are used in Industry 4.0 settings for real-time surveillance, predictive maintenance and self-regulation. The challenge in such environments is to provide reliable, low-latency and energy-efficient communication in the face of varying network topologies, limited node capabilities, and traffic dynamics. In this work, we present an edge-deployed digital twin framework with time-series-based predictive routing for Industrial IoT (IIoT) enabled WSNs. The suggested model uses a digital twin of the physical sensor network, hosted at the edge, to virtualize industrial resources and network conditions in real time, enabling local analysis and control. Time-series analysis is used to model temporal changes in sensor data, traffic patterns, and network links. This enables the routing protocol in the prediction of the network states and dynamically adjust the communication routes accordingly. The integration of edge computing and digital twin techniques curtails the reliance on cloud-based computing and capacitate near real-time processing at the network edge. The method enhances Packet Delivery Ratio (PDR), end-to-end delay, throughput and network lifetime performance over traditional routing mechanisms in IIoT WSNs. Through extensive simulations, it is demonstrated that the suggested approach performs better than traditional IIoT routing algorithms in the aforementioned parameters. These findings demonstrate the potential of edge computing, digital twins and time-series prediction analytics for delivering smart, resilient Industry 4.0 communication systems.