Target tracking in Internet of Things using reinforcement learning

The Internet of Things is a network of interconnected devices, sensors, and physical systems that communicate over the internet. These devices intelligently collect, process, and share data, enabling seamless interaction between the digital and physical worlds. IoT has broad applications in smart homes, healthcare, industrial automation, and smart cities. A key challenge in IoT environments is target tracking—identifying and monitoring moving objects within a defined area. Sensor nodes play a vital role in this process by detecting signals emitted from targets. Various algorithms are used for tracking, with differing levels of efficiency. Some rely on a single sensor node, which may result in low accuracy due to limited coverage. Others involve all nodes in collaborative tracking, which improves accuracy but significantly increases energy consumption due to excessive communication. A more balanced approach involves mobile sinks—movable nodes that follow the target within the environment. This paper proposes a novel method using reinforcement learning to train mobile sinks for more effective tracking. The proposed model enables adaptive movement and intelligent decision-making in the sink. Simulation results indicate that this approach improves detection accuracy, data delivery rates, and communication efficiency, making it a promising strategy for energy-aware and responsive IoT-based tracking systems.