STATISTICAL DESCRIPTION OF A CLASS OF NONGAUSSIAN CHANNELS USING EXPANSIONS BASED ON THE LOGNORMAL DISTRIBUTION.
Abstract
To develop improved detection procedures for the extraction of digital data from a background of nongaussian noise, it is necessary to obtain improved representations for a class of nongaussian processes. In this report a new representation theory based on the deviations from a lognormal distribution is developed. Two conceptually different representations are presented. These improved representations are particularly useful in obtaining a more complete description of the statistics for multi-hop and tropospheric scatter communication links. By the proper application of this new and flexible representation theory it is shown that one can develop improved detection procedures. The detectors based on the new representation can be adjusted to perform optimally for a large class of nongaussian channels. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 01, 1967
- Accession Number
- AD0660607
Entities
People
- Dirk R. Klose
- Ludwik Kurz
Organizations
- New York University