Parameter Selection for the Muntner-Wolf Model.
Final technical rept.,
ROME AIR DEVELOPMENT CENTER GRIFFISS AFB N Y
Pagination or Media Count:
This study is concerned with parameter assignment to the Muntner-Wolf model representing error sequence generation by a real channel. The model describes the distribution of the errors in time. A set of 124 troposcatter experiments from Project MALLARD is examined by a specifically developed set of computer routines. Twelve parameters, constants or simple equations, are found representative of all the experiments. An analytical expression for the central statistic used in parameter assignment is developed for the particular modulation system applied in the experiment. For measurable parameter values, it is found to depend only on signal-to-noise ratio and order of diversity. In order to determine the appropriate length of experiment for model building purposes, the experimental variance for the same statistic is developed. It is found to depend on the signal-to-noise ratio, the length of experiment and the RMS fading bandwidth. The entire problem of parameter error development is examined. A conclusion is reached that parameter assignment is practical using computer techniques.
- Non-Radio Communications