Validation of channel model for evaluating the performance of probing signal design in shallow tropical waters

Abstract

Active sensing has been the preferred technique for sediment classification to identify underwater objects and characterization of bottom type. The performance of the sediment classification techniques depends on the underwater channel fluctuations and the signal design of the probing signal. The multiple interaction of the probing signal, with the surface, volume and the bottom of the underwater channel and the random fluctuations of the medium conditions result in corresponding modifications of the probing signal. The distortions in the tropical shallow waters are known to be severe and directly impact the active sensing performance. The smart design of the probing signal can potentially improve the performance of the sediment classification techniques. Conventional signal design using Linear Frequency Modulation (LFM), has its limitations in the tropical shallow waters and alternative techniques need to be explored. Multiple combinations of frequency slopes for the two-stage LFM have been tried out for their effectiveness in overcoming the limitations of mismatch losses. The effectiveness of the signal design can be evaluated only based on the precise channel model that can accurately map the tropical shallow water environment where these systems will be deployed. The performance measures have been quantified using the Signal to Noise Ratio (SNR) of the Power Spectral Density (PSD) levels of the competing signals. An important contribution of this work is the validation of the simple Rogers model for shallow water channels used to explain the channel behaviour and its benchmarking with the rigorous BELLHOP channel model used for channel estimation. The simulation studies are a prelude to the deployment planning for an Autonomous Underwater Vehicle (AUV) deployed Digital Thin Line Array (DTLA) that can access very shallow depths and thus opens up significant application possibilities. Probing signal design, customized to the unique channel characteristics of the deployment site is the key focus of this paper. The DTLA with 12 array elements towed by an AUV that dives to 6 m depth in a shallow water area with depths as low as 25 m, has been used as an experimental scenario for the study. The transmitter is fitted on board the AUV and the DTLA sensors are towed behind with a towing cable.