Exact Performance of General Second-Order Processors for Gaussian Inputs
Abstract
The characteristic functions of general second-order processors with nonstationary nonzero mean Gaussian inputs is derived in closed form. Three classes of processors are considered: in the first, the decision variable is the sum of K independent terms, each of second-order form involving two statistically dependent Gaussian random variables; the second class of processor is a narrowband cross-correlator of arbitrary Gaussian processes, with accumulation of K independent low pass filter output samples; in the third class, the decision variable is a general quadratic-plus-linear form of M random variables, all statistically dependent on each other. Specializations to various forms of weighted energy detectors and correlators are made. Also, the characteristics function for the first class of processor subject to fading is evaluated. Programs for evaluating the cumulative and exceedance distribution functions of all three classes of processors are given and have been used to plot representative examples of performance. A comparison with a simulation result corroborates the analysis and program of the first class of processor.
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 15, 1983
- Accession Number
- ADA134098
Entities
People
- Albert H. Nuttall
Organizations
- Naval Underwater Systems Center