Underwater channel modelling, digital modem design and simulation

Abstract

Due to the unique propagation characteristics of acoustic waves and the presence of noise from natural and man-made sources in underwater environments, modems used for such purposes often require custom tuning for specific locations and use-case scenarios. This thesis delves into the impact of increased reflections in underwater environments on Underwater Acoustic (UWA) channels. Our investigation, conducted within a metallic shipyard in Singapore, uncovers amplified sparsity and multipath attributes. We suggest that modems with lower transmission rates are more suitable in scenarios with diverse and prolonged delays. Bhattacharya Distance results indicate a Nakagami distribution for channels with increased reflections. In the domain of underwater modem design, the thesis focuses on channel estimation in underwater Orthogonal Frequency Division Multiplexing (OFDM) communication. Utilizing Bellhop to simulate ideal increased-reflection scenarios, we explore Least Squares (LS), Minimum Mean Square Error (MMSE), and Orthogonal Matching Pursuit (OMP) channel estimators . Noise reduction algorithms are evaluated, with the DFT-based scheme and SVD algorithm enhancing estimator performance. The thesis introduces an upgraded Pilot-Assisted OFDM modem system with DFT-SVD noise reduction, effectively mitigating channel estimation noise in low Signal-to-Noise Ratio scenarios, thus offering valuable insights for enhancing the robustness and performance of UWA communication systems.