LZSS Output Encoding with Stream-Wise Compression for Better Ratios
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data compression algorithms, such as LZSS, do not achieve optimal compression ratios for all data types and often introduce additional redundancy, requiring multiple stages that increase processing time for both compression and decompression.
Innovation Solution
The method involves applying Lempel-Ziv-Storer-Szymanski (LZSS) processing to produce intermediate data streams, which are then grouped and further compressed using techniques like run-length encoding, adaptive arithmetic coding, and the Burrows-Wheeler Transform, allowing for improved compression ratios while maintaining efficient processing times.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a single LZSS compression algorithm is used, then the processing time is short, but the compression ratio is not optimal for all data types
Solution Approach 1:
The patent divides the input data stream into multiple segments and applies different compression algorithms to different segments. Specifically, it separates data into frequently-repeating patterns and less-repeating patterns, then applies optimized LZSS processing to each segment type, achieving high compression ratios across diverse data types without requiring a single complex multi-stage algorithm
Solution Approach 2:
The patent dynamically adjusts compression algorithm parameters based on data characteristics. It modifies the LZSS algorithm to use variable-length offset and length fields, and adjusts the search window size and matching threshold based on the statistical properties of the input data, allowing optimal compression performance across different data types while maintaining a single-pass processing approach
2Productivity
If multiple compression stages are used to improve compression ratio, then the data can be compressed further, but the processing time increases
Solution Approach 1:
The patent performs preliminary analysis of the data stream to identify frequently-repeating patterns before applying compression. By pre-processing the data to mark and categorize repeated sequences, the algorithm can then apply highly optimized compression to these identified patterns in a single pass, achieving multi-stage compression results without the iterative processing time
Solution Approach 2:
The patent extracts and separately processes frequently-repeating patterns from the data stream. It identifies these patterns first, extracts them into a separate processing stream, applies optimized compression specifically tailored for repeated sequences, then merges the results back, achieving high compression ratios while avoiding the need to process the entire data stream through multiple general-purpose compression stages
Data Source
AI summary
A data processing method where a source bit stream is manipulated to produce four intermediate streams: flag bits, literals, offsets and lengths. Flag bits are grouped into multi-bit units, and an output stream containing flag units, literals, offsets and lengths is emitted. The output stream occupies fewer bits than the source bit stream, but encodes all the data of the source bit stream.