Cycle frequency estimation of time/frequency diversity transmissions in a dense noise environment

A cycle rate estimation technique for weak diversity transmissions in noise is presented. The processing procedure is based on performing a weighted-average discrete Wigner distribution inner product operation at the input to a discrete Fourier analyzer followed by a low-frequency block then a weighted running sum across the fast Fourier transform (FFT) analyzer output. Fourier transformation across overlapping blocks of the summed outputs are then averaged to produce a cycle frequency estimate. The cycle frequency estimator is tested on a frequency hopped Welch 11 code Costas array in 0-dB SNR and on a dual-tone frequency hop/4-ary frequency shift keyed (FSK) signal where random time jitter and Gaussian pulse shaping have been employed.<
>