Lossless Data Compression with Wavelet Encoding for High I/O Speed

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing lossless compression algorithms struggle to achieve high compression speeds necessary for handling high data rates in scientific experiments, such as astroparticle physics, where data needs to be processed in real-time.

Innovation Solution

The method employs a computer-implemented approach using morphological wavelet transforms and bit range encoding to compress consecutive unsigned 16-bit integers, allowing for high parallelization and efficient use of AVX2 instructions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of substance

If general-purpose lossless compression algorithms like gzip are used, then good compression ratios are achieved, but compression speed is insufficient

Engineering Contradiction:
Improvecompression ratioVSAvoidcompression speed
Core Design Contradiction:
Loss of substanceVSProductivity

Solution Approach 1:

The patent divides the compression task into processing blocks of consecutive unsigned integers. Each block is handled independently through morphological wavelet transform and bit range encoding, enabling parallel processing and significantly improving compression speed while maintaining effective compression ratios for correlated scientific data

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent transforms the data representation by applying morphological wavelet transforms to convert consecutive integers into a domain where compression is more effective. It also dynamically determines bit ranges for encoding residuals, adapting the encoding parameters to the actual data characteristics to achieve both speed and compression efficiency

Inventive Principle:
Principle #35Parameter changes

2Productivity

If compression is applied to high data rate streams, then I/O speed is improved, but compression speed must be sufficiently high to handle the input data rate

Engineering Contradiction:
ImproveI/O speedVSAvoidcompression speed
Core Design Contradiction:
ProductivityVSSpeed

Solution Approach 1:

By segmenting the data stream into blocks and processing them independently with morphological wavelet transforms, the patent enables high-speed parallel compression that can keep up with high data rate inputs while reducing the compressed data volume to improve I/O speed

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent replaces traditional mechanical compression algorithms with a computational approach using morphological wavelet transforms and bit range encoding, leveraging modern CPU capabilities to achieve compression speeds sufficient for high data rate scientific data streams

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentEP4136755B1Method for fast lossless compression
Publication Date: 2025.04.02 MAX PLANCK GESELLSCHAFT ZUR FOERDERUNG DER WISSENSCHAFTEN EV
  • EP4136755B1 patent drawingFigure 1
  • EP4136755B1 patent drawingFigure 2
  • EP4136755B1 patent drawingFigure 3

AI summary

A computer-implemented method for compressing digital data, comprises the steps: - obtaining a sequence of digital data values; - mapping the sequence of digital data values to a sequence of code words having non-uniform bit lengths; - packing the sequence of code words into a sequence of storage words having uniform bit length; - outputting the sequence of storage words. According to the invention, the method further comprises the step of outputting information indicating the bit length of each code word.