Adaptive Entropy Coding with Grouped Literal Blocks

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Traditional adaptive entropy coding compression algorithms face inefficiencies due to the dispersion of literal symbols throughout the compressed data stream, leading to wasted compression opportunities and fractional bit loss when encoding symbols that do not match the current probability model.

Innovation Solution

Collecting literal symbols and compressing them together, either at the beginning or end of the stream, and using multi-level escape tokens to improve compression efficiency by reducing the number of bits required to encode literals and minimizing fractional bit loss.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of energy

If traditional adaptive entropy coding compression algorithms are used, then compression is achieved through statistical modeling, but literal symbols are dispersed throughout the compressed data stream causing wasted compression opportunities and fractional bit loss

Engineering Contradiction:

Improvecompression efficiencyVSAvoidfractional bit loss

Core Design Contradiction:

Loss of energyVSLoss of information

Solution Approach 1:

The patent segments the compressed data stream into two distinct parts: a code portion containing entropy-coded symbols and a separate literal portion containing collected literal symbols. This segmentation allows literals to be compressed together as a group rather than being dispersed individually throughout the stream, eliminating fractional bit loss and improving overall compression efficiency.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent merges all literal symbols into a single collective literal portion that is compressed together using adaptive entropy coding. By combining scattered literals into one group and applying compression algorithms to the entire set, the system recovers compression opportunities that would be lost if literals were encoded individually, thereby reducing fractional bit loss.

Inventive Principle:

Principle #5Merging (Combining)

2Productivity

If literals are encoded individually throughout the data stream, then encoding is simple, but compression opportunities are wasted and output data size increases

Engineering Contradiction:

Improvecompression ratioVSAvoidencoding complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The patent performs preliminary collection of all literal symbols during the compression process before final encoding. By gathering literals into a separate portion during the compression pass and then compressing them collectively in a subsequent stage, the system achieves better compression ratios without significantly increasing overall complexity, as the collection phase integrates naturally with the existing adaptive modeling process.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS7821426B2Adaptive entropy coding compression output formats

Publication Date: 2010.10.26 RED HAT INC

AI summary

A system stores compressed literal symbols in a first data block and encoded literal symbols in a second data block. The compressed literal symbols correspond to a first group of literal symbols and the encoded literal symbols correspond to a second group of literal symbols. Each of the second group of literal symbols occurs subsequently in a symbol stream to a literal symbol with the same value in the first group of literal symbols.