Boris I. Shakhtarin

Quasi-likelihood detection algorithm of the rectangular quasi radio signal with unknown duration

We considered the new quasi-likelihood detection algorithm of a wideband quasi radio signal with unknown amplitude, initial phase and duration against white Gaussian noise. We found the structure and statistical characteristics of the introduced detection algorithm and also we investigated the influence of the prior signal duration ignorance on detection efficiency.

Adaptive estimation of time and power parameters of the random pulse with inexactly known duration in the presence of the interference with unknown intensity

Synthesis of measurement algorithm of time delay, mathematical expectation and dispersion of the low-frequency random pulse with inexact known duration observed against white noise and band Gaussian interference, adapting for unknown intensities of action additive distortions is implemented. On the basis of the performed generalization of a local Markov approximation method the closed analytical expressions for conditional biases and variances of made estimates are found. By means of statistical computer modeling the functionality of the offered measurer is established and applicability borders of asymptotically exact formulas for characteristics of its operating effectiveness are determined.

Measurement of stepwise change point of the fast fluctuating Gaussian random process under conditions of the parametrical prior uncertainty

In this paper we propose a technically simple way of measurement of the abrupt change of fast fluctuating Gaussian signal under conditions of parametric prior uncertainty as an example of the mathematical expectation jumping of a random process with unknown intensity. Using local Markovian approximation of the solving statistics increments the definition technique of asymptotic characteristics of change-point time estimate is illustrated. Applying statistical computer modeling, we have found that measurers synthesized on the basis of the proposed approach are operable and the theoretical formulas describing their performance well conform to the corresponding experimental data in a wide range of parameter values of the analyzed process.

Measurer of the random radiopulse with free-form envelope and unknown time-and-frequency and power parameters

In the present work we consider the estimation algorithms of the appearance time and dispersion of the high-frequency Gaussian random pulse with free-form envelope and unknown band center against white noise. We suggested a quasi-optimal estimator of time-and-frequency and power pulse parameters, which verges towards performance of optimal analogues, but is technically more simple. For a calculation of estimation algorithms characteristics of signals with free-form envelope and unknown discontinuous parameters, the local Markov approximation method generalization was implemented. Conclusions and recommendations were confirmed by statistical computer modeling.

Threshold characteristics of the appearance time estimate of the random radio pulse with free-form envelope and inaccuracy unknown duration

In the present work we consider the estimation algorithm of the appearance time of the high-frequency Gaussian random pulse with free-form envelope against white noise. Contrary to the known studies, we supposed the useful signal duration to be known inaccurately. We found the asymptotically exact analytical dependences for the conditional bias and variance of the appearance time estimate on the basis of the local Markov approximation method. Then, by the methods of statistical computer modeling, we established that the received theoretical results adequately agrees with the corresponding experimental data in a wide range of output signal-to-noise ratios.