Underwater Acoustic Networks (UANs) hold enormous potential for both military and civilian applications. However, current networking protocols implemented are often sub-optimal, resulting in severely underutilized networks. We believe one of the key reasons for this shortcoming is a lack of good underwater acoustic propagation models for simulation packages used to evaluate UAN protocols. This thesis addresses this problem by developing a computationally-efficient approximation of a sophisticated analytical model called Monterey-Miami Parabolic Equation model (MMPE). The approximation can then be used to support UAN simulations. The characteristics of the problem make a statistical approach the methodology of choice for this study. Data was generated using the MMPE model. The data was used to develop a much less complex approximation for which an OpNet simulation module could be developed. The latter allows UAN operation to be modeled over a collection of nodes and over an interval of time, rather than a single point in time between two specific nodes, as modeled by the original equation. The result of this research can enable a more complete analysis of network enabling protocols and support more informed decisions regarding the appropriate node topology and protocols to use in order to increase network performance.

Approved for public release; distribution is unlimited.

http://archive.org/details/underwatercousti109451792