SE-PSI: Fog/Cloud server-aided enhanced secure and effective private set intersection on scalable datasets with Bloom...

Area of Science:

• Cryptography and Security
• Big Data Analytics
• Computer Science

Background:

• Private Set Intersection (PSI) is crucial for applications like credit evaluation and medical systems.
• Traditional PSI methods face performance and scalability challenges in the big data era.
• Existing protocols struggle to meet the demands of large-scale, privacy-preserving data analysis.

Purpose of the Study:

• To propose two novel, secure, and effective Private Set Intersection (SE-PSI) protocols.
• To address the performance and scalability limitations of traditional PSI for large datasets.
• To enhance privacy by hiding set/intersection size from the server in one protocol.

Main Methods:

• Developed two SE-PSI protocols utilizing deterministic encryption and Bloom Filters.
• Designed one protocol for high efficiency under a semi-honest server.
• Developed a second protocol for security against an economic-driven malicious server, concealing intersection size.

Main Results:

• Protocols achieved processing times of approximately 15 and 24 seconds for million-element datasets.
• A novel multi-round mechanism was introduced, nearly doubling efficiency.
• Experimental evaluation confirmed the high performance and scalability of the proposed SE-PSI protocols.

Conclusions:

• The proposed SE-PSI protocols offer efficient and scalable solutions for big data analysis.
• These protocols enhance security and privacy in set intersection computations.
• The multi-round mechanism provides a significant efficiency boost for practical applications.