The advent of 5G technology has markedly accelerated the development of the Industrial Internet of Things (IIoT), enabling faster and more reliable connectivity for various IoT devices. Many of these IIoT systems rely on sensor-based communication networks to monitor, collect, and transmit real-time data for industrial applications such as smart manufacturing, automated control systems, and predictive maintenance. However, this increased reliance on 5G networks introduces new cybersecurity risks, particularly the threat of rogue base stations that can intercept, manipulate, and disrupt data communications. In this study, we aim to identify and address the security threats posed by rogue base stations in the IIoT. Our detection approach is based on reference signal received power (RSRP) analysis, which allows us to monitor and evaluate signal strength variations. Since rogue base stations often transmit stronger signals to lure IIoT devices, abnormal fluctuations or inconsistencies in RSRP values can serve as key indicators of their presence. We use machine learning techniques, including recurrent neural networks, long short-term memory networks, and gated recurrent unit networks, to analyze and classify these signal patterns effectively. By leveraging the sequential nature of RSRP data, our model detects deviations from normal base station behavior, enabling the real-time identification of potential rogue base stations. By enhancing the security of sensor-driven IIoT systems, our approach ensures the protection of critical industrial operations that depend on real-time and accurate sensor data transmission.