Compressed Stream Conversion for Parallel Decompression Compatibility
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing compression methods, such as Deflate, require sequential extraction of variable-length codes, making parallelization of decompression difficult, and converting between different compression formats incurs high hardware costs.
Innovation Solution
A conversion device that includes a demultiplexer, extractors, and a deinterleave unit to reorder variable-length codes into a format compatible with common decompression methods, allowing parallel processing and reducing conversion costs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a first compression method generating interleaved substreams is used, then parallel decompression processing is enabled, but compatibility with external apparatuses using common compression methods deteriorates
Solution Approach 1:
The patent introduces a conversion unit as an intermediary component that transforms data from the first compression format (interleaved substreams) to the second compression format (common format like Deflate). This mediator enables compatibility with external apparatuses without sacrificing the parallel processing advantages of the first compression method, as the conversion occurs after decompression in the memory system.
2Adaptability or versatility
If recompression using a second compression method is performed for compatibility, then interoperability with external apparatuses is improved, but hardware cost and device complexity increase
Solution Approach 1:
The patent extracts only the necessary functionality for format conversion from a full recompression system. Instead of implementing a complete second compression method hardware, the conversion unit selectively transforms only the compressed data stream from the first format to the second format, removing unnecessary compression algorithms and keeping hardware costs low.
Solution Approach 2:
The conversion unit implements a lightweight, simplified data transformation process that avoids the complexity and high cost of full recompression hardware. By using a dedicated conversion component rather than a complete compression system, the patent achieves compatibility at minimal hardware investment.
3Measurement precision
If sequential extraction of variable-length codes is used, then decompression accuracy is maintained, but processing speed deteriorates
Solution Approach 1:
The patent segments the compressed data stream into multiple fixed-width substreams with interleaved variable-length codes. This segmentation allows parallel extraction of codes from different substreams simultaneously, maintaining decompression accuracy through proper reordering while dramatically increasing processing speed through parallelization.
Solution Approach 2:
The patent transforms the sequential structure of traditional compressed streams into a multi-dimensional interleaved structure where codes from different substreams can be processed in parallel. By reorganizing data in this alternative dimensionality, the system achieves both high-speed parallel processing and accurate decompression through proper reordering.
Data Source
AI summary
According to one embodiment, a conversion device includes a demultiplexer, first to Nth extractors and a deinterleave unit. The demultiplexer extracts first to Nth substreams from a first compressed stream. The first to Nth substreams are placed in order in the first compressed stream and include first variable-length codes to Nth variable-length codes into which first symbols to Nth symbols of a symbol string have been converted. The first to Nth extractors extract the first variable-length codes to the Nth variable-length codes from the first to Nth substreams. The deinterleave unit reorders the first variable-length codes to the Nth variable-length codes in accordance with the symbol string and outputs a second compressed stream.