Date of Award
Open Access Dissertation
Doctor of Philosophy in Electrical Engineering (PhD)
Administrative Home Department
Department of Electrical and Computer Engineering
Committee Member 1
Michael C. Roggemann
Committee Member 2
Committee Member 3
Influenced by environmental conditions, underwater acoustic (UWA) communication channels exhibit spatial and temporal variations, posing significant challenges for UWA networking and applications. This dissertation develops statistical signal processing approaches to model and predict variations of the channel and relevant environmental factors. Firstly, extensive field experiments are conducted in the Great Lakes region. Three types of the freshwater river/lake acoustic channels are characterized in the aspects of statistical channel variations and sound propagation loss, including stationary, mobile and under-ice acoustic channels. Statistical data analysis shows that relative to oceanic channels, freshwater river/lake channels have larger temporal coherence, higher correlation among densely distributed channel paths, and less sound absorption loss. Moreover, variations of the under-ice channels are less severe than those in open water in terms of multipath structure and Doppler effect. Based on the observed channel characteristics, insights on acoustic transceiver design are provided, and the following two works are developed. online modeling and prediction of slowly-varying channel parameters are investigated, by exploiting their inherent temporal correlation and correlation with water environment. The temporal evolution of the channel statistics is modeled as the summation of a time-varying environmental process, and a Markov latent process representing unknown or unmeasurable physical mechanisms. An algorithm is developed to recursively estimate the unknown model parameters and predict the channel parameter of interest. The above model and the recursive algorithm are further extended to the channel that exhibits periodic dynamics. The proposed models and algorithms are evaluated via extensive simulations and data sets from two shallow-water experiments. The experimental results reveal that the average channel-gain-to-noise-power ratio, the fast fading statistics, and the average delay spread can be well predicted. The inhomogeneity of the sound speed distribution is challenging for Autonomous underwater vehicles (AUVs) communications and acoustic signaling-based AUV localization due to the refraction effect. Based on the time-of-flight (TOF) measurements among the AUVs, a distributed and cooperative algorithm is developed for joint sound speed estimation and AUV tracking. The joint probability distribution of the time-of-flight (TOF) measurements, the sound speed parameters and the AUV locations are represented by a factor graph, based on which a Gaussian message passing algorithm is proposed after the linearization of nonlinear measurement models. Simulation results show that the AUV locations and the sound speed parameters can be tracked with satisfying accuracy. Moreover, significant localization improvement can be achieved when the sound speed stratification effect is taken into consideration.
Sun, Wensheng, "Online Learning of the Spatial-Temporal Channel Variation in Underwater Acoustic Communication Networks", Open Access Dissertation, Michigan Technological University, 2019.