An Intelligent Communication–Navigation Electronic System Based on Hierarchical Reinforcement Learning for Unmanned Surface Vehicles

Unmanned surface vehicles (USVs) play a vital role in maritime operations, yet they face significant challenges in maintaining reliable communication and accurate navigation in complex and dynamic electromagnetic environments. This paper presents a novel communication–navigation electronic system (CNES), designed to support mission-aware resource management and system adaptability under real-world marine conditions. The system incorporates an orthogonal chirp division multiplexing (OCDM) waveform and supports flexible task modes through dynamic subcarrier allocation mechanisms. To enhance the decision-making capability of CNES, we further propose an intelligent control algorithm, termed Integrated communication and navigation technology based on hierarchical reinforcement learning (ICNT-HRL). The ICNT-HRL algorithm adopts a two-layer control structure, where the high-level agent selects subcarrier block types based on mission modes, while the low-level agent jointly optimizes modulation schemes and power distribution. Simulation and field experiments are conducted by deploying ICNT-HRL on the CNES platform and benchmarking it against representative learning-based algorithms under identical conditions. Results demonstrate that ICNT-HRL achieves superior performance in terms of communication throughput, navigation accuracy, and system robustness, validating its practical effectiveness for real-world USV deployments.