Algorithm for Parallel Processing of a Speech Signal based on the Haar Wavelet

Digital signal processing is one of the fastest growing technologies today. Currently, it is developing more in the direction of artificial intelligence for the analysis and synthesis of speech signals. Digital signal processing is real-time informatics designed to solve the problems of receiving, processing, reducing redundancy and transmitting information at the rate of its receipt. And processing speed plays an important role in real-time speech signal processing. Timely collection, highspeed processing, and data transmission currently require highperformance hardware and software. But high-performance problems have not been fully resolved till present. This article is devoted to the analysis and research of methods for spectral analysis of a speech signal using the Haar wavelet, as well as parallel programming and development of a parallel processing algorithms based on the OpenMP and TBB standards. For hardware implementation, two multi-core processors with different characteristics were chosen.