Engineering topics

Korwar, V. N.

Publications and source records attributed to Korwar, V. N..

7 records

NASA NTRS1981

Coherent reference generator phase stability

Approximate phase stability estimates for the coherent reference generator (CRG) unit in the DSN's Frequency and Timing Subsystem (FTS) are calculated. The method used involves estimating the phase noise introduced by CRG components based upon measurements made in the past on similar components in other parts of the FTS and obtaining the CRG phase noise from the component phase noises. Three estimates of phase stability are calculated: the fractional frequency change for a 5 C step in temperature, the phase noise spectral density, and the Allan standard deviation. It is found from these estimates that the CRG phase stability is better than that of the H-maser physics unit + receiver. Thus, the first step in improving FTS phase stability would be to make improvements in the H-maser physics unit + receiver. These results are corroborated by indirect clock stability estimates calculated from Doppler data.

Korwar, V. N.↗

NASA NTRS1981

End-to-end quality measure for transmission of compressed imagery over a noisy coded channel

For the transmission of imagery at high data rates over large distances with limited power and system gain, it is usually necessary to compress the data before transmitting it over a noisy channel that uses channel coding to reduce the effect of noise introduced errors. Both compression and channel noise introduce distortion into the imagery. In order to design a communication link that provides adequate quality of received images, it is necessary first to define some suitable distortion measure that accounts for both these kinds of distortion and then to perform various tradeoffs to arrive at system parameter values that will provide a sufficiently low level of received image distortion. The overall mean square error is used as the distortion measure and a description of how to perform these tradeoffs are included.

Korwar, V. N.↗

NASA NTRS1981

Detection of gratings and small features in speckle imagery

The extent of picture degradation of speckle, in particular in synthetic aperture radar pictures, has been investigated in the cases where an observer has to detect (1) a small feature immersed in a darker background, and (2) a square wave grating. In each case, a theoretical model is developed for the observer's detection mechanism, and the probability of correct decision is related to relevant picture parameters such as contrast, looks per pixel, and size. These calculations are verified by psychophysical experiments using computer-simulated pictures. Detectability of gratings as a criterion for characterizing picture quality is shown to be far inferior to feature detectability.

Korwar, V. N.↗

NASA NTRS1981

Discrimination of form in images corrupted by speckle

The problem investigated is that of a human observer having to distinguish between certain specified geometrical forms corrupted by speckle - an idealization of the problem of a scientist studying a synthetic aperture radar map. Specifically, the cases of two simple alternative forms and of two and four orientations of a simple form have been considered. A theoretical model is developed for the observer's decision process by analogy with optimal receiver theory, and the probability of a correct decision is related to form parameters like size, contrast, and looks. These calculations are verified by psychophysical experiments using computer-simulated pictures.

Korwar, V. N.↗

NASA NTRS1978

Effect of pixel dimensions on SAR picture quality

In an SAR mapping system, the product of looks per pixel and number of pixels in the scene is kept constant. Assuming that the returns from all resolution cells obey Rayleigh statistics, the expression for pixel SNR incorporating both speckle and additive white Gaussian noise was derived. It is shown that it is possible to use fine resolution and leave the large-area estimate slightly but not much worse than if a larger pixel size had been initially decided upon.

Pierce, J. R.↗

NASA NTRS1978

Effect of ambiguities on SAR picture quality

The degradation of picture quality is studied for a high-resolution, large-swath SAR mapping system subjected to speckle, additive white Gaussian noise, and range and azimuthal ambiguities occurring because of the non-finite antenna pattern produced by a square aperture antenna. The effect of the azimuth antenna pattern was accounted for by calculating the aximuth ambiguity function. Range ambiguities were accounted for by adding appropriate pixels at a range separation corresponding to one pulse repetition period, but attenuated by the antenna pattern. A method of estimating the range defocussing effect which arises from the azimuth matched filter being a function of range is shown. The resulting simulated picture was compared with one degraded by speckle and noise but no ambiguities. It is concluded that azimuth ambiguities don't cause any noticeable degradation but range ambiguities might.

Korwar, V. N.↗

NASA NTRS1978

Effect of ambiguities on SAR picture quality

The degradation of picture quality in a high-resolution, large-swath SAR mapping system caused by speckle, additive white Gaussian noise and range and azimuthal ambiguities occurring because of the nonfinite antenna pattern produced by a square aperture antenna was studied and simulated. The effect of the azimuth antenna pattern was accounted for by calculating the azimuth ambiguity function. Range ambiguities were accounted for by adding, to each pixel of interest, appropriate pixels at a range separation corresponding to one pulse repetition period, but attenuated by the antenna pattern. It is concluded that azimuth ambiguities do not cause any noticeable degradation (for large time bandwidth product systems, at least) but range ambiguities might.

Korwar, V. N.↗