卷 60, 编号 2 (2017)

This is the third paper in a series of papers dedicated to the peculiarities of estimation of the continuous energy spectrums of random processes of different nature, which are determined by their samples at discrete moments of time. In the article we justify the methodology and present the quantitative results of analytical and experimental investigation and comparison of statistical characteristics of classical and “parametric” methods of energy spectrums reconstruction for interperiod fluctuations of different nature reflections (including the ones from meteorological objects) in pulse radars. The methodology is followed by quantitative results which correspond to and obtained for a real-world “adaptive” case. Under the latter, a priori unknown echoes’ covariance matrix is replaced with different-kind estimates formed from finite-size training samples. Based on the results obtained, we substantiate the spectral estimation algorithms reasonable for utilization in different-purpose radars, in particular in pulse Doppler weather ones. Discussion of efficient ways for their practical implementation on a unified basis of adaptive lattice filters concludes the paper.

This article considers the design of thin stripline (1 mm) bandpass filters in the centimeter band, which contain dielectrics with different relative permittivity ε_r. This includes the choice of the resonators, the dielectric material, the preliminary assessment of resonators’ unloaded quality factors and taking into account the features of the frequency response curve of the filter. It was found, that the passband width of the thin comb filters with λ/4 resonators and array-type filters with λ/2 resonators cannot exceed 6% for any values of ε_r and the length of resonators of at least 2 mm. The array-type filters with resonators of half-wave type and with alternating signs of the coupling factors between resonators were proposed. Under certain additional conditions, the attenuation poles appear in such filters, resulting in improved selectivity. The results of computer modelling of the frequency response curve of thin filters of cm band are presented. They are compared to the frequency response of other filters. The data obtained from computer modelling showed good correspondence with the experimental data.