Nested Cuckoo Hashing for Private Set Intersection Privacy

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing systems for calculating private set intersections in a multi-party setting are inefficient, resource-intensive, and leak information beyond the intersection, necessitating an improved hashing scheme for secure computation.

Innovation Solution

An extended private set intersection nested cuckoo hashing scheme is employed, involving the exchange of outer and inner hash functions to populate client and server hash tables, using cuckoo hashing to reduce comparisons and maintain data privacy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If hashing is used to improve PSI performance, then computation efficiency is improved, but information leakage about item placement positions increases

Engineering Contradiction:

Improvecomputation efficiencyVSAvoidinformation leakage

Core Design Contradiction:

ProductivityVSLoss of information

Solution Approach 1:

The patent divides the hash table into multiple buckets and uses multiple hash functions to map items to different buckets. This segmentation prevents direct observation of item placement positions while maintaining efficient lookup, as items are distributed across multiple segments rather than concentrated in single positions.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces an intermediary randomization mechanism where items are mapped through multiple hash functions to intermediate bucket positions before final placement. This intermediary layer obscures the direct relationship between item keys and their storage positions, preventing information leakage while preserving computational efficiency.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Reliability

If existing PSI protocols are used, then set intersection can be computed, but resource consumption and processing time increase

Engineering Contradiction:

Improveset intersection computationVSAvoidresource consumption

Core Design Contradiction:

ReliabilityVSUse of energy by moving object

Solution Approach 1:

The patent pre-computes and exchanges hash function sets between parties before the main computation. This preliminary action enables the actual PSI computation to proceed more efficiently using these pre-shared hash functions, reducing the resource consumption and processing time during the intersection calculation phase.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent uses a partial hashing approach where not all items need to be fully processed through all hash functions. The protocol can terminate early when intersections are found or when sufficient confidence is achieved, performing only the necessary computational actions rather than exhaustive processing, thus reducing overall resource consumption.

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentUS12470397B2Secure computation of shared data intersections

Publication Date: 2025.11.11 SAP SE

AI summary

Mechanisms are disclosed for providing an extended private set intersection nested cuckoo hashing scheme for secure computation of private set intersections. A set of outer hash functions and a set of inner hash functions are exchanged between a server and a client. A server outer hash table is populated using an outer hash function in the set of outer hash functions and the set of server items to create a set of server bins. For each server bin in the server outer hash table, server items to be inserted are iteratively placed into a server inner cuckoo hash table using a plurality of inner hash functions corresponding to each server bin, based on determining that a first corresponding hash table position is free.