Segmented Differential Compression for Offset-Misaligned Data Blocks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Standard data compression methods, including deduplication and differential compression, face inefficiencies in identifying and utilizing similar data across blocks with different offsets or unique data, leading to suboptimal compression ratios and increased decompression requirements.
Innovation Solution
The system divides input data blocks into segments, computes similarity degrees with other blocks, and selects optimal reference blocks for differential compression, allowing for improved compression efficiency by utilizing identical data spread across multiple reference blocks and reducing the number of blocks needed for decompression.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If standard differential compression is applied to entire data blocks, then compression is simplified, but compression ratios are suboptimal when blocks have different offsets or unique data
Solution Approach 1:
The patent divides input data blocks into multiple segments and computes similarity degrees for each segment independently against other data blocks. This allows different segments to reference different optimal reference blocks, thereby improving compression ratios by capturing partial similarities that would be missed in whole-block comparison, while maintaining the simplicity of differential compression through modular segment-based processing
2Productivity
If variable size deduplication is used to improve compression ratios, then compression efficiency improves, but device complexity increases
Solution Approach 1:
The patent applies segmentation to divide data blocks into segments, enabling variable-size reference matching without requiring variable-size block structures. This achieves compression efficiency similar to variable size deduplication by allowing segments of different lengths to reference optimal blocks, while maintaining simpler fixed-block processing architecture, thus reducing system complexity
Solution Approach 2:
The patent introduces a new dimension of segment-level similarity computation independent of block size. Instead of changing block sizes (variable size deduplication), it operates at the segment level within fixed blocks, achieving similar compression efficiency through a different dimensional approach that simplifies system architecture
3Productivity
If multiple reference blocks are selected for each segment, then compression efficiency improves, but computational overhead increases
Solution Approach 1:
The patent computes similarity degrees for each segment against multiple candidate blocks but selects only the single optimal reference block for each segment based on the highest similarity degree. This applies partial action by avoiding full differential compression computation for all candidate blocks, achieving compression efficiency through selective optimal matching while reducing computational overhead by limiting processing to the best match
Data Source
AI summary
A method for differential compression includes receiving input data blocks that are selected for compression. For each input data block, the input data block is divided into at least two segments. For each of the at least two segments, a similarity degree between the respective segment and each of the data blocks excluding the respective data block is computed. For each of the at least two segments, the data block which has a biggest similarity degree with the respective segment among the data blocks excluding the respective data block is selected as an optimal reference data block for the respective segment. The differential compression is applied to the input data block and optimal reference blocks in response to determining a differential compression that is to be applied based on the similarity degree between the segments of the input data block and the corresponding optimal reference blocks.