Encrypted Search Latency via Mobile Transit Processing

The patent segments the encrypted search process into two distinct phases: an offline preprocessing phase where the encrypted dataset is processed to generate searchable indexes and data structures, and an online query phase where actual search operations are performed. This segmentation allows computationally intensive operations to be performed in advance, significantly reducing the time required for actual search queries while maintaining security through encrypted data structures.

The patent performs preliminary actions by pre-processing the encrypted dataset to generate optimized search indexes, inverted indexes, and other data structures before queries are executed. This preliminary processing creates efficient lookup structures that enable fast query execution without requiring decryption or intensive computation during the actual search operation, thereby reducing latency while preserving security.

The patent divides the large encrypted dataset into manageable segments and processes each segment independently to generate localized indexes and data structures. This segmentation allows the system to handle large datasets by breaking them into smaller processing units, reducing the computational burden on any single processing operation while maintaining the ability to search across the entire dataset comprehensively.

The patent replaces traditional mechanical search mechanisms that require decryption and linear scanning of large datasets with cryptographic-based search structures such as encrypted inverted indexes and trapdoor functions. This substitution enables efficient searching of large encrypted datasets using mathematical operations rather than computationally expensive decryption and sequential scanning processes.