Underwater Acoustic Signal Processing Workshop

Listings Underwater Acoustic Communications: Working at the Intersection of Physics, Signal Processing, and Communications Theory James C. Preisig Woods Hole Oceanographic Institution WHIO, MS #11 Woods Hole, MA 02543 [email protected] The underwater environment is widely regarded as one of the most difficult communication channels. Underwater acoustic communications systems are challenged by the characteristics of acoustic propagation through the underwater environment in which they operate. There are a wide range of physical processes that impact underwater acoustic communications and the relative importance of these processes are different in different environments. The characteristics of the environment can impact everything from the determination of the optimal topology for distributing network nodes to efficient transmit signal selection and receiver algorithm design. While a large number of the pertinent questions regarding these issues are still open research topics, research over the past decade has certainly lent insights into their answers. A particularly interesting and challenging environment is that in which surface scattered signals form a significant portion of the received signal. Rapid channel fluctuations and a extended delay spread of the channel can combine to make it difficult for coherent communications algorithms to track the channel accurately enough to enable reliable high-rate communications. The rapid channel fluctuations limit the averaging time that most simple algorithms can employ to estimate important channel parameters which often results in ill-condition estimation or adaptation problems. Effective techniques for addressing this problem include slowing down the apparent rate of channel fluctuation or reducing the number of free parameters that must be adjusted to effectively track the channel. The consideration and/or exploitation of relevant channel physics can be used to realize improvements using these techniques as demonstrated with several examples in this talk. [This work is supported by ONR Grants #N00014-05-10085, #N00014-06-10788, #N00014-07-10184, and #N00014-07-10738 and NSF Grant #OCE-0519903.]