Underwater acoustic channel modeling: methods, tools, and development trends

Objectives The objective is to support performance evaluation of underwater acoustic communication algorithms under different propagation conditions. Methods Existing underwater acoustic communication channel modeling methods are reviewed and classified as deterministic or stochastic, and related simulation tools are compared in terms of their applicability and uses in communication assessment. Results Direct replay models can be used to retest algorithms under measured channel conditions, models based on empirical relations can be used for link parameter analysis, and models based on acoustic propagation theory can be used for communication assessment under specified environmental and transmitter-receiver conditions. Geometry-based stochastic models can be used to examine the effects of changes in geometry and motion conditions on communication performance, while statistical stochastic models and generative models support repeated stochastic evaluations under the corresponding channel conditions based on prescribed statistical characteristics and learned data distributions, respectively. On this basis, HARBOR, an underwater acoustic channel simulation platform, is proposed. The platform integrates measured-channel replay, multiple channel models, and configurable Doppler, noise, and interference effects to support communication performance evaluation under different channel conditions and multiple influencing factors. Conclusions The applicability of existing models is constrained by the completeness of environmental information and the coverage of measured data. Model selection should therefore account for the requirements of the specific communication evaluation task. Future research should focus on quantitatively determining the applicability limits of channel models, expanding measured-channel test sets across multiple scenarios, and standardizing reference channel models for representative propagation conditions.