Diversity in Shallow Water Environments Using Blind Time-Frequency Separation Techniques

The shallow water environment can be characterized as a time-dispersive system whose time-varying impulse response can be expressed as a superposition of time-frequency components with dispersive structures. In this paper, a blind time-frequency processing technique is employed to separate these components without knowledge of environmental parameters. This technique is based on first approximating the time-frequency structures of the received signal, and then designing separation filters based on time-frequency warping techniques. Based on this method, a receiver is developed to exploit the diversity of the channel and to improve communications performance.