Federated Learning for Intelligent Sensing and Navigation Technologies for 6G Communications

Integrated sensing and communication (ISC) is a crucial technology for sixth‐generation (6G) networks, enabling the real‐time sensing and transmission of data for latency‐sensitive applications. However, device heterogeneity and asynchronous sensing intervals pose severe threats to resource allocation and data availability. This paper proposes a directed intelligent sensing scheme (DISS) that integrates it with federated learning (FL) to address the asynchronous data sensing and inefficient resource allocation in 6G ISC networks. DISS provides concurrent sensing periods, dynamic resource support, and distributed model training without compromising data privacy. FL detects unmonitored sensing demands and optimizes device‐side resources to enhance responsiveness and data availability. Performance metrics, including latency, data availability, and resource usage, are monitored and optimized in real time through the intelligent scheduling of sensing periods. The simulation results validate that DISS is superior compared to the current approaches, and data availability and resource use are improved by 13.72% and 12.09%, respectively. Such advancements enable applications such as autonomous driving and UAVs in extremely dynamic 6G scenarios.