The Adaptive Antenna Processing in Pseudo-Noise Communication System.
Shang-Min Juang · Deep Blue (University of Michigan) · 1982
The output signal to noise ratio of adaptive antenna processing is limited by the value of the product of the number of antenna elements and the signal to non-directive ambient noise ratio. The output signal to noise ratio of a Pseudo-Noise (PN) signal system is restricted by errors made in predicted spatial noise distribution when the system was designed. The application of adaptive antenna processing in the PN system removes these limitations and provides high output signal to noise ratio over noise condition that consists of both ambient and spatial noises. There are two places at which to position the adaptive antenna processing in PN system. The output signal to noise ratios and the processing gains of these two configurations were derived analytically. Numerical values were calculated under different noise b and width and power levels for the one jamming noise condition. A new method, the noise power filter method, combining the parameters of the adaptive antenna processing and the spatial noises, is invented and applied to the performance derivation. This new method is powerful in h and ling the output noise power calculation of any system that uses the adaptive antenna processing. Real time operation algorithms were derived. The derivations were based on the following discoveries; (1) the independence between the adaptive output signal to noise ratio and the effect of signal power mixing in the noise covariance matrix; and (2) the magnitude of the weighting factor is inversely proportional to the signal power mixed in the noise covariance matrix. The former discovery is contrary to the current belief and because of this, no special device is required to separate the signal from the noise when estimating the noise covariance matrix. Because of the second discovery the algorithms are based only on the deleting of the weighting factor variation due to the variation of the mixed signal power in the noise covariance matrix. Three methods for decreasing the probability of error in peaked signal detection for the combined processing are suggested in this thesis. Finally computer simulation of the real time processing are included. The results indicate good agreement between the analytical derivation and the r and om simulations.