Accession Number : ADD018270

Title :   Adaptive Parameter Kernel Processor.

Descriptive Note : Patent, Filed 7 Aug 95, patented 27 Aug 96,

Corporate Author : DEPARTMENT OF THE NAVY WASHINGTON DC

Personal Author(s) : Bond, James W ; Schlosser, Thomas W

Report Date : 27 Aug 1996

Pagination or Media Count : 8

Abstract : The adaptive parameter kernel processor of the present invention uses an adaptive locally optimum detection (ALOD) signal processing algorithm to suppress interfering signals in a communications signal. The adaptive parameter kernel processor generates a signal magnitude value for each signal sample, partitions the signal magnitude values into sets, and concurrently executes multiple copies of the ALOD signal processing algorithm. Each copy of the signal processing algorithm applies a probability density function to the sets of signal magnitude values, generating a gain factor for each set. The probability density function includes a parameter that has a different value in each copy of the signal processing algorithm. Each copy of the signal processing algorithm generates an average gain factor associated with the corresponding parameter value. The parameter values are updated respectively in the multiple copies of the signal processing algorithm by a parameter comparator to cause the parameter value associated with the minimum average gain factor to approach the midpoint of the parameter values. An output signal is formed from the signal samples and an arithmetic function of the average gain factor associated with the midpoint of the parameter values.

Descriptors :   *SIGNAL PROCESSING, *RADIO RECEIVERS, *PATENTS, ALGORITHMS, OPTIMIZATION, PROCESSING EQUIPMENT, PROBABILITY DENSITY FUNCTIONS, KERNEL FUNCTIONS, GAIN, ADAPTIVE SYSTEMS

Subject Categories : Cybernetics
      Radio Communications

Distribution Statement : APPROVED FOR PUBLIC RELEASE