Progress on signal-selective direction finding

Abstract

The recently discovered Cyclic MUSIC algorithm for narrowband signal-selective direction finding using antenna arrays circumvents many drawbacks of conventional techniques by exploiting known spectral correlation properties (namely known cycle frequencies such as the baud rate or carrier frequency) of the desired signals to reject undesired signals, interference, and noise. Two recent advances in the capabilities of Cyclic MUSIC are described. The first enables Cyclic MUSIC to simultaneously estimate the directions of arrival of signals having different cycle frequencies instead of having to sequentially process each separate frequency in a list of cycle frequencies (either known a priori or measured). The second advance reduces the sensitivity of Cyclic MUSIC to error in the knowledge of the cycle frequency of interest by estimating the frequency of a quadratically-regenerated sine wave and then using that estimate as the cycle frequency parameter in the computation of the cyclic autocorrelation matrix, which is then processed to estimate the directions of arrival.<>