Publications
    - Reset

Defensive Publications are a method for the open source community to prevent the issuance of low quality software patents. These publications serve as evidence to disclose to the examiner what is new and innovative within the technology. Through Linux Defenders, developers collaborate with our team to ensure that these publications are effective in protecting the freedom to create, invent and operate.

Publication Date Title Tags
April 16, 2013 Optimizing Rolling Hash computation using SIMD Vector registers
Described is a technique for speeding up the rolling hash computation used in content-aware chunking for data deduplication via SIMD vector instructions.
single instruction multiple data vectorized hash function X86 Processor Vector Instructions streaming SIMD Extensions advanced vector expressions sliding window rolling hash content-aware chunking data de-duplication
April 16, 2013 Scalable Data Deduplication using Similarity Matching and in-memory indexes
Data Deduplication for large datasets typically encounters the problem of very large indexes to hold the chunk hashes. The indexes do not fit into RAM and require multiple disk accesses to lookup a hash to check whether it is already present in the archive.
data de-duplication
March 12, 2013 Optimized Min-heap based Similarity Detection for Delta Encoding
Described is a new technique for detecting approximate similarity between different chunks of data which enables the use of Delta Encoding techniques to reduce data volume, in addition to Deduplication based on exact matches.
data de-duplication MinHash delta encoding