A Numerical Simulator for an Autonomous, Bottom-mounted Sonar for Measurement of Mid-Frequency Reverberation

While bottom reverberation is typically dominant over surface reverberation at frequencies of a few kilohertz in shallow water, sea surface roughness has a significant impact on transmission loss and hence on reverberation. Measurements of this effect can be difficult with a system deployed from a moving ship due to both the system motion and the constraints of operating the ship safely in the rough seas of interest. To overcome these difficulties, the Autonomous Reverberation Measurement System (ARMS) has been developed and will be introduced as the first part of this presentation. This system is a benthic lander with a directional source and receive array mounted on a rotation stage. Powered by batteries and programmed prior to deployment, the system can measure reverberation from 2-6 kHz as a function of direction autonomously for up to 3 months. This system was deployed in the spring of 2017 as part of an experiment conducted off Geoje Island, Republic of Korea. During this experiment, the ARMS was deployed in 28 m of water and measured reverberation in muddy sand environment with a range-dependent water depth that was punctuated with rock outcroppings on the seafloor. Sample results from ARMS during the experiment will be shown.In order to validate propagation and reverberation models, a simulator for the ARMS is being developed. The basic idea behind the simulator is to generate numerical, time-domain data mimicking those measured by the system. To this end, the simulator needs to take environmental parameters such as sound speed field, rough sea surface and bottom, etc. as input. It also needs to be able to allow practical source waveforms as input. Finally, it needs to consider ARMS's actual source and receiver beam patterns as a function of frequency. As a first step, a rough-interface reverberation model is being developed for range-dependent environments. First-order perturbation theory is employed, and the unperturbed background medium can be layered and heterogeneous with arbitrary range dependence. To calculate the reverberation field, two-way forward scatter due to the slowly changing unperturbed environment is handled by fast numerical methods. Backscatter due to small roughness superimposed on any of the slowly varying interfaces is handled efficiently using a Monte Carlo approach.Although the ARMS is a monostatic system, its sensor geometry, however, makes it a slightly bi-static system. Therefore, the simulator needs to take the bi-static nature into considerations. Two approaches are adopted. The first one is a genuine bi-static simulation, where a 2-D realization of the environment is first generated, then reverberation time series for each of the receivers due to a pulsed source is calculated. The second one simulates beamed output, so a pseudo-monostatic method can be used. To increase numerical efficiency, several approximation techniques are employed. Of particular importance is the use of spatial windowing which effectively reduces the number of frequencies to be simulated for the purpose of Fourier synthesis. Numerical examples for several shallow water scenarios are presented to demonstrate the application of the simulator.