Predictive Data Compression for Lossless Low-Complexity Encoding

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing data compression techniques often result in loss of data and are inefficient in terms of computing time and cost, particularly when dealing with limited bandwidth or storage capacity.

Innovation Solution

A method that extrapolates data samples based on preceding values, differentiates them to remove redundancy, and compresses the differentiated samples without loss, using an algorithm of order N for linear complexity, which is implemented in a device capable of low-cost silicon chip integration.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If conventional data compression techniques are used, then data transmission or storage capacity is improved, but data loss occurs and computing time increases

Engineering Contradiction:

Improvedata transmission capacityVSAvoiddata loss

Core Design Contradiction:

Quantity of substanceVSLoss of information

Solution Approach 1:

The patent applies preliminary action by performing extrapolation of data samples before compression. The method extrapolates each sample based on preceding samples to create predicted values, then computes differences between actual and predicted values. This preliminary processing transforms the data into a form that reveals redundancies while preserving all information, enabling subsequent lossless compression.

Inventive Principle:

Principle #10Preliminary action

2Quantity of substance

If conventional data compression techniques are used, then data transmission or storage capacity is improved, but computing time becomes excessive

Engineering Contradiction:

Improvedata transmission capacityVSAvoidcomputing time

Core Design Contradiction:

Quantity of substanceVSLoss of time

Solution Approach 1:

The patent extracts the essential information from data samples by computing differences between actual samples and extrapolated predictions. Instead of compressing the raw samples directly, the method extracts only the deviation information, which contains the redundant elements that can be efficiently compressed while maintaining linear computational complexity.

Inventive Principle:

Principle #2Taking out (Extraction)

3Loss of information

If complex compression algorithms are used to achieve lossless compression, then data loss is eliminated, but device complexity and implementation cost increase

Engineering Contradiction:

Improvedata lossVSAvoidimplementation cost

Core Design Contradiction:

Loss of informationVSDevice complexity

Solution Approach 1:

The patent replaces complex mechanical or computational compression systems with a mathematical substitution approach. By using extrapolation based on preceding samples and computing simple differences, the method substitutes elaborate compression algorithms with straightforward arithmetic operations that are computationally efficient and easy to implement in hardware.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentUS10382055B2Digital data compression

Publication Date: 2019.08.13 COMMISSARIAT A LENERGIE ATOMIQUE ET AUX ENERGIES ALTERNATIVES

AI summary

A method for compressing digital data, including: extrapolating a value of each sample of data to be compressed as a function of a value of at least one preceding sample, to produce an extrapolated sample; differentiating between each extrapolated sample and the corresponding sample of data to be compressed, to produce a differentiated sample; and deleting redundancy between successive differentiated samples produced by the differentiating stage.