Channel models are crucial for theoretical analysis, performance evaluation, and deployment of wireless communication systems. Traditional channel sounding systems are insufficient for handling the dynamic changes of channels in the next-generation space–air–ground–sea integrated networks (SAGSIN), which often results in outdated channel models that fail to provide reliable prior information for communication systems. To address this challenge, this article proposes an integrated channel sounding and communication (ICSC) method as a practical solution. Unlike orthogonal frequency division multiplexing (OFDM), affine frequency division multiplexing (AFDM) provides a full delay-Doppler representation of the channel, achieving optimal diversity in time-frequency doubly dispersive channels and effectively addressing the aforementioned challenges. Thus, this article investigates the fundamental principles of AFDM, showing how it enables simultaneous communication and channel sounding, and explores key performance metrics for both functionalities. Additionally, the distinction and relationship between channel sounding, estimation, tracking, and scatterer sensing are clarified. Several potential application scenarios for AFDM-ICSC are also examined. Finally, the article highlights the key challenges in implementing AFDM-ICSC, outlines future research directions, and offers valuable insights for the continued development of this technology.
As wireless communication systems evolve toward intelligent operation, the growing complexity of dynamic and non-stationary propagation environments inherent in large-scale and heterogeneous Internet of Things (IoT) scenarios imposes stringent demands on link adaptation robustness. To address these challenges, we propose a joint bandwidth and modulation and coding scheme (MCS) optimization framework that leverages autonomous scenario identification (ASI) and dueling double deep Q-network (D3QN), named as ASI-D3QN. Specifically, a lightweight deep convolutional neural network (CNN) is tailored for ASI to achieve high identification accuracy while maintaining low computational complexity. This design is particularly suited for resource-constrained devices. Then, a D3QN-based optimization strategy is developed to seamlessly integrate ASI-derived environmental context with intrinsic channel metrics. The proposed framework explicitly expands the decision space by integrating signal bandwidth as an additional optimization dimension, facilitating adaptive optimization over multi-dimensional parameters. Finally, an intelligent communication prototype is established to comprehensively validate the proposed approach under representative standardized channel models. Experimental results demonstrate that, compared to existing methods, the proposed ASI achieves at least a 0.27% accuracy improvement with fewer model parameters. Moreover, relative to conventional link adaptation schemes, the ASI-D3QN strategy achieves superior throughput gains in complex dynamic environments.