Encrypted Segment Deduplication Using Hash-Based Identification

The patent divides data into fixed-size segments before encryption. Each segment is independently encrypted and stored. This segmentation allows the system to process and compare individual segments using their cryptographic hashes, enabling deduplication to work efficiently on encrypted data by comparing hash values rather than plaintext content.

The patent introduces cryptographic hash functions as an intermediary mechanism. Instead of directly comparing encrypted segments or plaintext segments, the system computes hashes of segments (which can be done on encrypted data) and compares these hash values. This intermediary approach allows identification of identical segments without exposing plaintext content, thus maintaining both security and deduplication efficiency.

The patent performs deduplication identification (through hashing) before compression is applied to the final storage format. By establishing segment identities through cryptographic hashing at the segmentation stage, the system creates a reference framework that allows later compression operations to proceed without losing the ability to track and identify identical segments through their pre-established hash values.

The patent uses cryptographic hash values as an intermediary identifier that remains valid through compression operations. While the actual segment data may be compressed, the hash value serves as a stable identifier that allows the system to recognize identical segments regardless of their compressed state, enabling deduplication to function effectively alongside compression.

The patent applies segmentation as the first processing step, dividing data into fixed-size segments before any encryption or compression operations. This initial segmentation creates discrete units that can be independently hashed and tracked. The segmented structure persists through subsequent encryption and compression, allowing the system to maintain segment boundaries and identify identical segments through their hash values even when the actual data undergoes multiple transformations.

The patent employs cryptographic hash functions as a robust intermediary that can operate on and identify segments regardless of their encryption or compression state. The hash computation serves as a bridge between the transformed data (encrypted and compressed) and the deduplication logic, allowing identical segments to be recognized through their hash values without requiring access to or knowledge of the plaintext content or the specific transformations applied.