Том 69, № 8 (2024)

The main stages of receiving and processing a linear frequency modulation (LFM) signal reflected from a point moving target are considered. When calculating the phase components of the received LFM signal, the principle of time scale transformation is used, which manifests itself in the mutual movement of the target and the radar station (radar) with pulsed radiation. At all stages of processing the LFM signal, including frequency transfer, quadrature demodulation (selection of the complex envelope of the signal) and matched filtering, the amplitude and phase functions of the signal are determined, allowing coherent processing of a batch of radio pulses. The analysis of the phase structure of the LFM signal is carried out, as a result of which the main phase components are determined, which are taken into account during interperiod signal processing in high-resolution radar.

An algorithm for spatial-temporal signal processing in discrete time is obtained. The estimation-correlation and estimation-correlation-compensation options for constructing optimal algorithms are considered, which also makes it possible to generalize to non-Markov signals and interferences. It is shown that in order to implement the obtained algorithms, it is necessary to use extrapolation one-step estimates of the signal and interference. It is established that the interference compensator is part of the optimal estimation-correlation-compensation algorithm. The application of the proposed approach makes it possible to obtain a variety of quasi-optimal algorithms for signal detection and parameter estimation. The resulting structures are convenient for implementation in digital form by hardware and software. A spatial-temporal estimation-correlation algorithm for processing the atmospheric signal has been obtained, which allows estimating the distance to a thunderstorm discharge taking into account the height of the observation point. The influence of the number of field components on the measurement error of the distance to the lightning discharge is investigated. To improve the accuracy of distance determination, it is proposed to use a three-component field model.

It is shown that the solution to the problem of ensuring stable and high-precision tracking of aero-dynamic targets of various classes can be obtained by synthesizing an algorithm for their complex tracking and recognition, taking into account their characteristic features based on the results of observation by heterogeneous information sources. A set of interrelated algorithms has been developed, including an algorithm for complex adaptive filtering of the phase coordinates of the observed target, conditional on its dynamic and signal signs obtained from observations of a set of sources; an algorithm for recognizing the class of the observed target based on the results of evaluating the values of its dynamic and signal signs, as well as the conditions for evaluating phase coordinates.

An approach to the construction of on-board consumer equipment for global navigation satellite systems as part of the on-board equipment of unmanned aerial vehicles (UAVs) is considered. The analysis of the features of the navigation support of UAVs in the process of implementing various scenarios of their application is carried out. The Markov theory of optimal estimation of random processes has been used to obtain algorithms for complex signal processing of a satellite radio navigation system and a 5G positioning module for the case of structurally separate processing of radio signals. A block diagram of the system is presented, a characteristic feature of which is the absence of cross-links between the channels of radio signal processing.

An algorithm for stabilizing the false alarm rate is proposed, based on the approximation of the probability distribution density of the decisive statistics counts during multi-frame accumulation against the background of noise described by the Rayleigh and Weibull distributions. It is shown that the proposed algorithm for stabilizing the false alarm rate is adaptive and invariant to the noise distribution law.

The replacement of a distributed radar object by a simple two-point matrix simulator is considered. In this case, the distributed object is replaced by a model composed of two points. It is shown that this leads to modeling errors, expressed in a change in the mathematical expected value of angular coordinate noise. Relationships have been obtained that allow, firstly, to estimate these errors and, secondly, to set the spacing of model points at which their acceptable level is guaranteed.