Application of response surface methodology to radar data compression

A compression technique has been developed by The Ohio State University. This technique analyzes the scattering centers of a return on an individual basis in the frequency domain where the return has a simple and smooth shape. Therefore, if these scattering centers are extracted and their individual responses found, each scattering center can be sampled fewer times due to its smooth shape and still be reconstructed accurately. In fact the technique has compressed image data by a factor as large as several hundred, depending upon the specific object. This study examines the many tradeoffs and assumptions of the compression and reconstruction codes through the application of response surface methodology (RSM). The goal was to identify key variables and use this information to expand the application of this code to a wide variety of platforms. Inherent in this study was the characterization of input variables as they relate to the reconstruction accuracy. The RSM also affords the ability to predict optimum settings for the variables. Thus, even inexperienced operators ideally would have a systematic method of choosing variables and settings to attain optimal results.