Search results

The analysis of electric energy quality with a required by modern standards accuracy demands applying of digital signals processing. At measurement, one of the main operations is finding signals correlation. Operations performance over digital representations of the signals containing a large points number requires many of computing resources that results in importance of a fast correlation calculation problem. The well-known methods of correlation calculation is applying a range of algorithms: transformation correlation to cyclic convolution, calculation cyclic convolution through FFT and others. These algorithms are well-studied, however their combination make possible additional optimization of algorithms. In paper we offer the method of computing procedure optimization for signals correlation calculation based on operations reduction when performing a sequence two Fourier transformations. Using another way of placing the calculations results we can reduce a data rearranging operations number. Depending on computing architecture, the offered method allows to accelerate calculations to 20 - 25%.