Reusing codebooks in parameter quantization

• US 20060080090A1
• Filed: 10/07/2004
• Published: 04/13/2006
• Est. Priority Date: 10/07/2004
• Status: Abandoned Application

• Current Assignee: Nokia Solutions and Networks US LLC (Nokia Corporation)
• Original Assignee: Nokia Corporation
• Inventors: Ramo, Anssi, Himanen, Sakari, Nurminen, Jani
• Application Number:
  US10/961,471
• Publication Number:
  US 20060080090A1
• Time in Patent Office: Days
• US Class Current: 704/222
• CPC Class Codes: G10L 19/07 Line spectrum pair [LSP] vocoders

Abstract

The present invention provides a new methodology for reusing codebooks for a multistage vector quantization of parameter quantizers of signals. Prior art multistage vector quantization is done in such a way that each stage has different optimized codebooks. The prior art codebooks, thus, use quite a lot of a memory storage space. Using the same codebook stages several times, according to the present invention, reduces the memory usage and a codebook structure maintains good quality by using optimized codebooks for the most important (first) stages in the quantization. The number of codebooks is reduced by reusing the same codebooks in the refining stages. Additionally, according to the present invention, using many predictors is space-wise efficient since they need only a few of coefficients instead of larger codebooks.

Claims

1. A method of reusing codebooks for a multistage vector quantization of parameter quantizers for a signal, comprising the steps of:

training multistage vector quantization codebooks for all predictor and non-predictor modes of said parameter quantizers;

analyzing said trained codebooks for different stages of said vector quantization and optionally analyzing corresponding training data used for said training and identifying similar codebooks corresponding to different predictor and non-predictor modes out of said all predictor and non-predictor modes for said different stages based on said analyzing using a predetermined criterion;

combining said training data corresponding to N codebooks selected from said similar codebooks based on a further predetermined criterion; and

training said N codebooks using said combined training data thus generating a new common codebook to be used instead of said N codebooks for said multistage vector quantization of said parameter quantizers for said signal, wherein N is an integer of at least a value of two.

2. The method of claim 1, wherein the step of said training multistage vector quantization codebooks also include training predictors corresponding to said all predictor modes of said parameter quantizers.

3. The method of claim 1, wherein the steps of said analyzing said combining and said training are repeated until a pre-selected level of memory space savings is reached.

4. The method of claim 1, wherein said N codebooks have the same size.

5. The method of claim 1, wherein said identifying similar codebooks using said predetermined criterion is based on evaluating a variance of related parameters, and optionally on evaluating the variance of training vectors or code vectors, corresponding to said similar codebooks.

6. The method of claim 1, wherein the step of said analyzing said trained codebooks includes evaluating at least one related parameter for an original codebook out of said trained codebooks for one predictor mode of said all predictor modes, and then evaluating said at least one related parameter using a different trained codebook out of said trained codebooks for a different predictor mode of said all predictor modes in place of said original trained codebook and using identical data for said both evaluatings.

7. The method of claim 6, wherein the step of said combining said training data include combining said training data for said original codebook and said different codebook if said predetermined criterion is met.

8. The method of claim 1, wherein said parameter quantizers contain both vector and scalar parameters.

9. The method of claim 1, wherein said training said N codebooks using said combined training data is performed using a pre-selected algorithm, optionally a generalized Lloyd algorithm.

10. The method of claim 1, wherein all steps are performed by an encoder of a communication system, and said encoder optionally is a part of a mobile device which is optionally a mobile phone.

11. The method of claim 10, said encoder is capable of storing said common codebooks and capable of generating an encoded quantized signal from said signal by using and reusing said common codebook for said multistage vector quantization of said parameter quantizers for said signal.

12. A computer program product comprising:

a computer readable storage structure embodying computer program code thereon for execution by a computer processor with said computer program code characterized in that it includes instructions for performing the steps of the method of claim 10 indicated as being performed by any component or a combination of components of said encoder.

13. An encoder capable of reusing codebooks for a multistage vector quantization of parameter quantizers for a signal, comprising:

means for training multistage vector quantization codebooks for all predictor and non-predictor modes of said parameter quantizers;

an analyzing block, for analyzing said trained codebooks for different stages of said vector quantization and optionally analyzing corresponding training data used for said training and identifying similar codebooks corresponding to different predictor and non-predictor modes out of said all predictor and non-predictor modes for said different stages based on said analyzing using a predetermined criterion; and

a combining block, for combining said training data corresponding to N codebooks selected from said similar codebooks based on a further predetermined criterion; and

means for training said N codebooks using said combined training data thus generating a new common codebook to be used instead of said N codebooks for said multistage vector quantization of said parameter quantizers for said signal, wherein N is an integer of at least a value of two.

14. The encoder of claim 13, wherein said training multistage vector quantization codebooks also include training predictors corresponding to said all predictor modes of said parameter quantizers.

15. The encoder of claim 13, wherein said analyzing said trained codebooks, said combining said training data and said training said N codebooks are repeated until a pre-selected level of memory space savings is reached.

16. The encoder of claim 13, wherein said N codebooks have the same size.

17. The encoder of claim 13, wherein said identifying similar codebooks using said predetermined criterion is based on evaluating a variance of related parameters, and optionally on evaluating the variance of training vectors corresponding to said similar codebooks.

18. The encoder of claim 13, wherein said analyzing said trained codebooks includes evaluating at least one related parameter for an original codebook out of said trained codebooks for one predictor mode of said all predictor modes, and then evaluating said at least one related parameter using a different trained codebook out of said trained codebooks for a different predictor mode of said all predictor modes in place of said original trained codebook and using identical data for said both evaluatings.

19. The encoder of claim 18, wherein said combining said training data include combining said training data for said original codebook and said different codebook if said predetermined criterion is met.

20. The encoder of claim 13, wherein said parameter quantizers contain both vector and scalar parameters.

21. The encoder of claim 13, wherein said encoder is a part of a communication system or a part of a mobile device which is optionally a mobile phone.

22. The encoder of claim 13, wherein said means for training said multistage vector quantization codebooks and said means for training said N codebooks using said combined training data are incorporated in one block.

23. The encoder of claim 13, further comprising:

a memory, for storing said common codebook; and

a coding module, capable of retrieving said common codebook from said memory for generating an encoded quantized signal from said signal by using and reusing said common codebook for said multistage vector quantization of said parameter quantizers for said signal.