Audio Signal Compression Using Adaptive Autocorrelation Correction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing audio compression methods face inefficiencies due to ill-conditioned cases caused by special input signals, leading to low compression efficiency and poor quality of reconstructed speech signals, as all signals are processed uniformly without considering their characteristics.

Innovation Solution

A signal compression method that adjusts autocorrelation coefficient correction factors based on the characteristics of each input signal, using techniques such as energy parameter calculation and lag-window adjustment, to produce more suitable linear prediction coefficients for accurate compression.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If all signals are processed in the same way using fixed autocorrelation coefficient correction factors, then the processing is simple and uniform, but ill-conditioned cases occur for special input signals leading to unstable solving and low compression efficiency

Engineering Contradiction:
Improveprocessing simplicityVSAvoidstability of autocorrelation matrix solving
Core Design Contradiction:
Ease of operationVSReliability

Solution Approach 1:

The patent applies dynamics by making the autocorrelation coefficient correction factors adaptive rather than fixed. The white-noise correction factor and lag-window are dynamically adjusted based on the energy parameter and zero-crossing rate of each input signal frame, allowing the processing to adapt to different signal characteristics and avoid ill-conditioned cases while maintaining operational simplicity through automated adaptation.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The patent changes parameters by modifying the autocorrelation coefficient correction factors based on signal characteristics. Specifically, the white-noise correction factor win_lag(0) and lag-window win_lag(k) are adjusted according to the energy parameter E and zero-crossing rate nz, transforming fixed parameters into variable ones that optimize stability and compression efficiency for different input signals.

Inventive Principle:
Principle #35Parameter changes

2Device complexity

If all signals are processed in the same way, then the processing method is simple, but the compression efficiency of lossless coder and quality of reconstructed speech signals are low

Engineering Contradiction:
Improveprocessing method complexityVSAvoidcompression efficiency
Core Design Contradiction:
Device complexityVSProductivity

Solution Approach 1:

The patent improves compression efficiency by changing parameters adaptively. The autocorrelation coefficient correction factors are modified based on signal-specific parameters (energy and zero-crossing rate), allowing the system to optimize compression for each signal type without requiring complex manual configuration, thus improving productivity while keeping the processing method relatively simple.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The system applies self-service by automatically calculating the energy parameter and zero-crossing rate for each signal frame and using these to self-adjust the autocorrelation coefficient correction factors. This eliminates the need for external manual tuning and allows the processing method to automatically adapt to different signal characteristics, improving compression efficiency without proportionally increasing complexity.

Inventive Principle:
Principle #25Self-service

3Ease of operation

If uniform processing is applied to all signals, then the processing is straightforward, but the quality of reconstructed speech signals deteriorates due to ill-conditioned cases

Engineering Contradiction:
Improveprocessing straightforwardnessVSAvoidquality of reconstructed speech signals
Core Design Contradiction:
Ease of operationVSManufacturing precision

Solution Approach 1:

The patent resolves this contradiction by changing parameters adaptively based on signal characteristics. The white-noise correction factor and lag-window are adjusted according to the energy parameter and zero-crossing rate, which improves the quality of reconstructed speech signals by preventing ill-conditioned cases while maintaining straightforward automated processing without manual intervention.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The system implements feedback by calculating signal characteristics (energy and zero-crossing rate) and using these to adjust the autocorrelation coefficient correction factors. This feedback mechanism ensures that the processing adapts to each signal's properties, improving reconstruction quality while keeping the operation straightforward through automated closed-loop control.

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP2204797B1Signal compression method and apparatus
Publication Date: 2011.12.14 HUAWEI TECH CO LTD
  • EP2204797B1 patent drawingFigure 1
  • EP2204797B1 patent drawingFigure 2
  • EP2204797B1 patent drawingFigure 3

AI summary

The present invention relates to audio compression. A signal compression method and apparatus are provided. The signal compression method includes: multiplying an input signal by a window function; calculating original autocorrelation coefficients of a windowed input signal; calculating a white-noise correction factor or a lag-window according to the original autocorrelation coefficients, and calculating modified autocorrelation coefficients according to the original autocorrelation coefficients, the white-noise correction factor and the lag-window; calculating linear prediction coefficients according to the modified autocorrelation coefficients; and outputting a coded bit stream according to the linear prediction coefficients. The technical solution under the present invention avoids ill-conditioned cases of special input signals, makes the modified autocorrelation coefficients more suitable for subsequent compression, improves the compression efficiency of a lossless coder and the quality of reconstructed speech signals of a lossy coder, and involves only simple operations.