The Strand Compression Algorithm

In order to examine critical data at a distant location, a requirement has been established for the transport of ALTAIR radar data from the collection site to a remote site (CONUS). The Strand Compression Algorithm described in this proposal is designed to take advantage of several redundant components inherent in the data: 1. Radar returns are processed intra-spatially using a predictor to decorrelate information over a localized region. 2. Radar returns are processed on an intra-scan basis using a predictor to decorrelate information between scans. 3. The data representation (code domain) is compressed to minimize data representation. This aspect relies on the fact that the return information will seldom use the entire dynamic range of the receiver. 4. Localized return bandwidth is evaluated and decimation is applied depending on the detected bandwidth. This concept relies on the fact that the radar returns do not continually use the entire receiver bandwidth. The algorithm is estimated to achieve an average efficiency of 18:1. This represents a factor of 6.0-9.0 times improvement over published lossless techniques. Data compression is of vital interest in the military and commercial communications community. The primary benefit of the Strand Compression Technique is the introduction of new "tools" that can be incorporated into a broad range of current and emerging algorithms. The techniques described will apply to arbitrary data, but will achieve the greatest advantage for band-limited data types including speech samples and imagery.