Performance characteristics of acoustic direction finders of localized broadband signal sources

Abstract

In this paper, we substantiate main performance characteristics of the regression algorithm of a two-channel direction finder of localized sources of broadband signal against the background of interference distributed in space, invariant to radiated power and supply voltage. We analyze mutual statistical characteristics of broadband signals in a two-channel direction finder. Based on the method of canonical decomposition in complex form, we show the cross-correlation function of processes at the outputs of direction finder antennas when operating on a localized signal source against the background of interference distributed in space. For invariance of performance characteristics, we consider the application of sign transformation to signals and analyze the dependence of a cross-correlation function of the input signals after the sign transformation at different signal-to-noise ratios. We show the formation of a directivity pattern in a correlation direction finder with signal-to-noise ratio more than 1. For direction-finding of localized signal source, we substantiated a regression algorithm with discrete-analog signal processing, where sign processing based on standard digital hardware components is used. The invariance of the algorithm to supply voltage is shown by representing Boolean functions as expressions of conjunction, disjunction, and negation. The regression algorithm is substantiated in the time domain. We obtain the expressions for mean and variance of the calculated statistics for direction-finding of localized signal source against the background of interference distributed in space. We show the possibility of formation of a directivity pattern with completely suppressed side lobes. The algorithm functions at signalto-noise ratios up to 1. The algorithms can be used in passive acoustic and hydroacoustic direction finders, as well as direction finders with ultra-wideband signals with computer-based signal processing.