Abstract
Acoustic waves are the preferred medium for long-range underwater communications. Increasing number of innovative methods for underwater communication using acoustic waves appears. Performance of underwater communications for the new methods needs to be evaluated through simulation. Simulation of underwater acoustic communications is challenging due to many impediments, including attenuation, multipath propagation, noise and Doppler spread. In this paper, a baseline time domain simulation model is extended to several frequency-domain models. The proposed frequency models, including two incoherent models and a coherent model, take multipath attenuation and ambient noise into account. An incoherent linear fitting model and a coherent model are simulated and compared with a theoretical reference and the baseline time model. The proposed incoherent models are also compared with one another. Simulation shows that the incoherent linear fitting model produces results similar to the multi-frequency-merge-path model, but requires less computation time. In addition, the proposed coherent model is compared with field experimental data. The coherent model with color noise, in the frequency domain, can match closely the bit error rates of the field experimental data.
Graphical Abstract
In this paper, several models are proposed and compared considering multipath attenuation and ambient noise. The models in the time domain and frequency domain all take into account the attenuation calculation along with white and colored noise. In the time domain, a classic baseline model is presented. In the frequency domain, not only incoherent but also coherent attenuation are proposed and implemented. Incoherent means that the interference created by these phase differences is ignored. Coherent refers that the phase differences of signals propagating through different paths are taken into account. White noise is modeled as traditional Additive white Gaussian noise. Colored noise makes the white noise go through a low-pass filter with a profile similar to the one of the underwater ambient noise. The bit error rate versus energy per bit to noise power spectral density ratio Eb/N0 is applied for evaluation. The simulations are implemented using MATLAB and the BELLHOP tracing program.
Acknowledgements
We would like to thank Brian S. Borowski for sharing his MATLAB® code with us. We acknowledge financial support from Natural Sciences and Engineering Research Council (NSERC) of Canada (Engage Grant).
Notes
No potential conflict of interest was reported by the authors.