Fast Frequency-Wavenumber Analysis and Fisher Signal Detection in Real-Time Infrasonic Array Data Processing

A high speed algorithm for computation of frequency-wavenuniber (f-k) spectra is developed, and two real-time infrasonic data processing techniques that it makes possible, are described: (1) Signal detection by a search of f-k space. This process is compared to the N-4 correlator, a broad-band signal detector. The f-k search with a Fisher detector has a theoretical advantage, which we verify in practice. (2) An f-k filter technique for calculating ' best beam ' estimates. This technique traces the beam containing maximum power, from frequency to frequency through f-k space, and thus allows for wandering of signal velocity and arrival azimuth. This maximum power function is taken as the frequency spectrum of the best beam. In our programs the Fisher statistic of the signal estimate, and the velocity and azimuth, are computed and displayed as functions of frequency. Examples from real data for both processing techniques are discussed.