A Correlation Polarimeter for Noise-Like Signals

Optimum estimation (tracking) of the polarization plane of a linearly polarized electromagnetic wave is determined when the signal is a narrow-band Gaussian random process with a polarization plane angle which is also a Gaussian random process. This model is Compared to previous work and is applicable to space communication. The estimator performs a correlation operation similar to an amplitude -comparison monopulse angle tracker, giving the name correlation polarimeter. Under large signal-to-noise ratio (SNR), the estimator is causal. Performance of the causal correlation polarimeter is evaluated for arbitrary SNR. Optimum precorrelation filtering is determined. With low SNR, the performance of this system is far better than that of previously developed systems. Practical implementation is discussed. A scheme is given to reduce the effect of linearly polarized noise.     

On Signal Fluctuations in Tracking Systems

All tracking systems suffer, to some extent, from signal level fluctuations. The signal fluctuations are modeled by a Gaussian process and the noise performance of both linear and nonlinear first-order tracking systems is considered. In the nonlinear case, the Ito versus Stratonovich calculus difficulty arises, but it is resolved to make sense physically. Several analytical results are obtained which are useful in analyses of tracking systems.     

Optimum Position Estimation Using Multiple-Frequency Omega

The maximum-likelihood estimator of the line of position (LOP) resulting from K frequencies received from each of two Omega stations for high SNR is derived. The estimator takes advantage of interfrequency correlation of phase uncertainties by modeling them as resulting from phase velocity uncertainties which are jointly Gaussian with known covariance. The utilization of Omega for search and rescue systems is examined. For this application, the result of increasing the size of the unambiguous lanes of Omega by addition of a fourth frequency to the present three-frequency format is examined. The results of Monte Carlo simulation, performed for several easily implemented fourth frequencies, are presented. The simulation results illustrate the extent to which lane errors can be expected as a function of various fourth frequencies, phase velocity uncertainties, and SNR.     

Digital Radiometer Performance

This work investigates the performance of an all-digital, total-power radiometer (TPR), consisting of RF, mixer, and IF stages followed by an analog-to-digital converter (ADC). Square-law detection and smoothing is performed in a digital computer. A figure of merit, the degradation factor, is defined which compares the digital TPR with an ideal analog TPR. Exact results are obtained which include the effects of saturation, finite step size, and finite sampling rate of the ADC. A good approximation is developed which compares favorably with the exact results under certain conditions. The use of a general digital filter for smoothing is considered. A new parameter, the ?equivalent summation number?, is shown to be analogous to the equivalent integration time of an analog TPR.