Quasioptimal Estimates of the Center Frequency of a Narrow-Band Normal Random Process with a Gaussian Spectrum
V. S. Sobolev, F. A. Zhuravel'
Keywords: random processes, center frequency of the narrow-band spectrum of a normal random process, most plausible estimates, Doppler measuring systems, model experiment
Pages: 71-77
Abstract
Based on the well-known algorithm of the most likely estimates of the center frequencies of a normal narrow-band random process with a fractional rational spectrum, a similar algorithm for a process with a Gaussian spectrum is proposed and investigated by computer simulation. The practical importance of the development of algorithms for efficiently estimating the parameters of such processes is that they are a sufficiently close model of the output signal of laser Doppler systems used to study hydrodynamic and aerodynamic flows, measure the velocity of diffusely reflecting solid objects, and perform laser sensing of the atmosphere. The results of the studies show that the proposed algorithm does not shift estimates of the Doppler frequency (velocity), gives fairly accurate estimates (for a Doppler system with 25 interference fringes in the measuring volume at a measurement time equal to 100 Doppler periods, the error is only 0.12%), has high noise immunity, and evaluates the resulting errors.
|
