Iterative Rate-Distortion Optimization for AAC Encoding

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional Advanced Audio Coding (AAC) encoding algorithms are limited in optimizing rate-distortion performance due to the separate optimization of quantization step sizes and Huffman codebooks, which can lead to non-convergent solutions and neglect the potential of quantized spectral coefficients as optimization parameters.

Innovation Solution

An iterative rate-distortion optimization algorithm using Lagrangian multipliers is employed to jointly optimize scale factors, Huffman codebooks, and quantized spectral coefficient sequences, ensuring full decoder compatibility and minimizing a cost function that balances rate and distortion.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If separate optimization of quantization step sizes and Huffman codebooks is used, then encoding simplicity is maintained, but rate-distortion performance is limited and may not converge

Engineering Contradiction:
Improverate-distortion optimization performanceVSAvoidencoding algorithm complexity
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The patent combines separate optimization of quantization step sizes and Huffman codebooks into a unified joint optimization framework. The Lagrangian cost function integrates both parameters along with quantized spectral coefficients, allowing simultaneous optimization that achieves better rate-distortion performance while ensuring convergence through the unified mathematical framework.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The patent introduces an iterative optimization algorithm that dynamically adjusts quantization step sizes, Huffman codebook selections, and quantized spectral coefficients across multiple iterations. The algorithm uses Lagrangian multipliers that are updated iteratively to converge to optimal values, transforming static separate optimization into dynamic joint optimization.

Inventive Principle:
Principle #15Dynamics

2Productivity

If quantized spectral coefficients are excluded from optimization parameters, then encoding complexity is reduced, but compression efficiency is lost

Engineering Contradiction:
Improvecompression efficiencyVSAvoidoptimization parameter complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent merges quantized spectral coefficients into the optimization parameter set alongside quantization step sizes and Huffman codebooks. The Lagrangian cost function treats all three parameters equally, enabling joint optimization that fully exploits compression opportunities while maintaining a unified mathematical framework that ensures convergence.

Inventive Principle:
Principle #5Merging (Combining)

3Manufacturing precision

If iterative joint optimization is applied, then rate-distortion tradeoff is improved, but computational complexity increases

Engineering Contradiction:
Improverate-distortion optimization performanceVSAvoidcomputational complexity
Core Design Contradiction:
Manufacturing precisionVSPower

Solution Approach 1:

The patent performs preliminary actions by pre-defining the Lagrangian cost function structure and selecting appropriate Huffman codebooks from a fixed set before iterative optimization begins. This preliminary setup reduces the computational burden during iteration by limiting the search space and pre-organizing optimization parameters.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The iterative optimization algorithm uses feedback mechanisms where the Lagrangian multiplier λ is updated based on the distortion and bit rate from previous iterations. This feedback loop allows the algorithm to converge to optimal values by continuously adjusting parameters based on performance measurements, ensuring both optimality and convergence.

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP2346031B1Rate-distortion optimization for advanced audio coding
Publication Date: 2015.09.30 BLACKBERRY LTD
  • EP2346031B1 patent drawingFigure 1
  • EP2346031B1 patent drawingFigure 2
  • EP2346031B1 patent drawingFigure 3~4

AI summary

A method for optimization of rate-distortion for Advanced Audio Coding (AAC). The method provides for the identification of quantized spectral coefficient sequences for optimization of rate-distortion. The method also provides joint optimization of scale factors, Huffman codebooks and quantized spectral coefficient sequences for minimization of a rate-distortion cost. The method provides an iterative rate-distortion optimization algorithm for AAC encoding. In each iteration, the method first finds the optimal scale factors and quantized spectral coefficients when Huffman codebooks are fixed, then updates Huffman codebooks and quantized spectral coefficients given the optimized scale factors. The iterations may be applied until a predetermined threshold is attained.