Parallel signal processing system and method

• US 10,127,925 B1
• Filed: 06/26/2017
• Issued: 11/13/2018
• Est. Priority Date: 12/17/2007
• Status: Active Grant

• Current Assignee: Patent Armory Inc. (Jeffrey M. Gross)
• Original Assignee: Calltrol Corporation
• Inventors: Wu, Wai
• Primary Examiner(s): King, Simon
• Application Number:
  US15/633,211
• Time in Patent Office: 505 Days
• CPC Class Codes: G10L 15/34 Adaptation of a single recogniser for parallel processing, e.g. by use of multiple processors or cloud computing
• CPC Class Codes: G10L 25/18 the extracted parameters being spectral information of each sub-band
• CPC Class Codes: G10L 25/21 the extracted parameters being power information
• CPC Class Codes: G10L 25/27 characterised by the analysis technique
• CPC Class Codes: H04M 19/02 providing ringing current or supervisory tones, e.g. dialling tone or busy tone
• CPC Class Codes: H04M 7/129 Details of providing call progress tones or announcements

Abstract

A system and method for processing a plurality of channels, for example audio channels, in parallel is provided. For example, a plurality of telephony channels are processed in order to detect and respond to call progress tones. The channels may be processed according to a common transform algorithm. Advantageously, a massively parallel architecture is employed, in which operations on many channels are synchronized, to achieve a high efficiency parallel processing environment. The parallel processor may be situated on a data bus, separate from a main general purpose processor, or integrated with the processor in a common board or integrated device. All, or a portion of a speech processing algorithm may also be performed in a massively parallel manner.

Claims

1. A method for processing audio signals, comprising:

(a) storing a representation of at least one audio stream in at least one memory;

(b) storing a common instruction sequence in the at least one memory;

(c) controlling at least one single-instruction, multiple-data type parallel processor having a plurality of processing cores, to process the at least one audio stream according to the common instruction sequence across the plurality of the processing cores, to perform at least one of frequency transform domain processing, wavelet transform processing, and time-to-frequency domain transform processing, wherein portions of the at least one audio stream are concurrently and synchronously processed according to the common instruction sequence across the plurality of the processing cores, to produce at least one processed audio stream having different audio information than the representation of the at least one audio stream; and

(d) storing the at least one processed audio stream in the at least one memory.

2. The method according to claim 1, wherein the at least one audio stream comprises a plurality of audio streams, and the at least one processed audio stream comprises a composite of the plurality of audio streams.

3. The method according to claim 1, wherein the at least one audio stream comprises a plurality of audio streams, wherein the common instruction sequence comprises instructions for performing echo processing.

4. The method according to claim 1, wherein the at least one audio stream comprises a plurality of audio streams, wherein the common instruction sequence comprises instructions for performing Fourier transform processing.

5. The method according to claim 1, wherein the at least one audio stream comprises a plurality of audio streams, wherein the common instruction sequence comprises instructions for performing both time domain transform processing and frequency domain transform processing of the audio information in the plurality of audio streams.

6. The method according to claim 1, wherein the single-instruction, multiple-data type parallel processor comprises a multiprocessor having a common instruction decode unit for the plurality of processing cores, each processing core having a respective arithmetic logic unit, all arithmetic logic units within a respective multiprocessor being adapted to simultaneously execute the common instruction sequence, wherein the single-instruction, multiple-data type parallel processor executes under control of a general purpose processor.

7. The method according to claim 1, wherein the single-instruction, multiple-data type parallel processor is a graphic processing unit of a computer system, which interfaces with the central processing unit through a Peripheral Component Interconnect Express (PCIe) interface.

8. The method according to claim 1, wherein the common instruction sequence comprises instructions for automatically extracting instructions from the at least one audio stream.

9. The method according to claim 1, wherein the common instruction sequence comprises instructions for automatically performing at least one speech recognition primitive.

10. The method according to claim 1, wherein the common instruction sequence comprises instructions for automatically identifying at least one in-band audio signal.

11. A system for processing audio band signals, comprising:

(a) an input configured to receive a representation of at least one audio stream;

(b) at least one memory configured to store information derived from the representation of the at least one audio stream; and

(c) at least one single-instruction, multiple-data type parallel processor having a plurality of processing cores, configured to;

read the information derived from the representation of the at least one audio stream stored in the at least one memory;

process the information derived from the representation of the at least one audio stream according to a common instruction sequence, across the plurality of the processing cores, to perform at least one type of processing selected from the group consisting of frequency domain transform processing, wavelet domain transform processing, and time-to-frequency domain transform transform processing, wherein portions of the at least one audio stream are concurrently and synchronously processed according to a common instruction sequence across the plurality of the processing cores;

generate at least one processed audio stream dependent on the type of processing having different audio information than the representation of the at least one audio stream; and

store the at least one processed audio stream in the at least one memory.

12. The system according to claim 11, wherein the at least one audio stream comprises a plurality of audio streams, wherein the at least one processed audio stream represents contributions from each of the plurality of audio streams.

13. The system according to claim 11, wherein the at least one audio stream comprises a plurality of independent audio streams, and the at least one processed audio stream comprises a plurality of output audio streams, corresponding to the plurality of independent audio streams.

14. The system according to claim 11, wherein the common instruction sequence comprises at least one of instructions for performing echo processing and instructions for performing Fourier transform processing.

15. The system according to claim 11, wherein the single-instruction, multiple-data type parallel processor comprises a multiprocessor having a common instruction decode unit for the plurality of processing cores, each processing core having a respective arithmetic logic unit, all arithmetic logic units within a respective multiprocessor being adapted to simultaneously execute the common instruction sequence, wherein the single-instruction, multiple-data type parallel processor executes under control of a general purpose processor.

16. The system according to claim 15, wherein the single-instruction, multiple-data type parallel processor is a graphic processing unit of a computer system, which interfaces with the central processing unit through a Peripheral Component Interconnect Express (PCIe) interface.

17. The system according to claim 11, wherein the common instruction sequence comprises instructions for automatically extracting instruction information from the audio stream, the at least one single-instruction, multiple-data type parallel processor being further configured to selectively process the information derived from the representation of the at least one audio stream according to the extracted instruction information, and store the extracted instruction information in the at least one memory.

18. A method for processing audio signals, comprising:

(a) receiving audio stream information;

(b) automatically performing at least one process selected from the group consisting of a time-to-frequency domain transform process and a wavelet domain process on the received audio stream information, with at least one single-instruction, multiple-data type parallel processor having a plurality of processing cores, wherein portions of the audio stream information are concurrently and synchronously transformed according to a common instruction sequence across the plurality of the processing cores; and

(d) communicating a processing result of the process on the received audio stream information through a digital communication interface.

19. The method according to claim 18, wherein the audio stream information comprises a plurality of audio streams, wherein the processing result represents contributions from each of the plurality of audio streams.

20. The method according to claim 18, wherein the audio stream information comprises information derived from a plurality of independent audio streams, and the processing result comprises a plurality of independently processed independent audio streams.