Multi-Server Data Reconciliation via Multi-Party Computation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing systems fail to securely reconcile impression and conversion data while maintaining user privacy and preventing fraudulent activity, as they are vulnerable to data interception and manipulation.

Innovation Solution

Implementing a multi-party computation (MPC) process using two aggregation servers to decrypt doubly-encrypted impression and conversion data, employing blind signatures and commutative encryption to verify data integrity and prevent fraud, while ensuring user anonymity through k-anonymity and differential privacy techniques.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If data is encrypted using a single public key, then decryption is simplified, but user privacy is compromised and tracking is enabled

Engineering Contradiction:
Improveuser privacy protectionVSAvoidencryption complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent divides the encryption task into multiple independent parts by using multiple public keys (first public key and second public key) instead of a single key. Each key encrypts a portion of the data, and both encrypted portions are required for complete decryption. This segmentation prevents any single entity from accessing the full plaintext data, thereby protecting user privacy while maintaining a manageable encryption structure.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent implements nested encryption by encrypting the impression data with the first public key and then encrypting the result with the second public key (or vice versa). This creates a layered encryption structure where the data is protected by multiple encryption layers. To decrypt, the process must reverse through each layer using the corresponding private keys, ensuring that no single party can access the data without contributing their cryptographic credential.

Inventive Principle:
Principle #7Nested doll (Nesting)

2Reliability

If multiple aggregation servers are used, then fraud prevention is improved, but system complexity increases

Engineering Contradiction:
Improvefraud prevention capabilityVSAvoidsystem complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent combines multiple aggregation servers (first aggregation server and second aggregation server) into a unified system that processes data together. Each server receives data from client devices and performs computations, but the results are aggregated and verified across both servers. This merging approach allows the system to leverage multiple independent entities to cross-validate data and prevent fraud while sharing the computational load.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The patent introduces an intermediary verification mechanism where the first aggregation server and second aggregation server mutually verify each other's data processing. Each server acts as an intermediary that not only processes data but also validates the work of the other server. This intermediary relationship ensures that neither server can unilaterally manipulate data, as both must agree on the results, thereby preventing fraud while distributing system complexity across multiple coordinated components.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Measurement precision

If impression and conversion data are reconciled, then attribution accuracy is improved, but user tracking is enabled

Engineering Contradiction:
Improveattribution accuracyVSAvoiduser tracking
Core Design Contradiction:
Measurement precisionVSObject-affected harmful factors

Solution Approach 1:

The patent uses cryptographic hashing to create a copy of the user identifier that can be used for matching impression and conversion data without revealing the actual user identity. The hash function creates a fixed-length string that is computationally infeasible to reverse, allowing the system to match records based on the hash value while the original user identifier remains concealed. This copying approach enables accurate attribution measurement while preventing user tracking.

Inventive Principle:
Principle #26Copying

Solution Approach 2:

The patent transforms the user identifier parameter into a hashed version before using it for data reconciliation. This parameter change maintains the functional capability of matching impression and conversion records (through the hash value) while eliminating the ability to track or identify the original user. The hash function preserves the equality relationship between identifiers while obfuscating the actual user information, achieving attribution accuracy without user tracking.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP3895365B1Preventing data manipulation using multiple aggregation servers
Publication Date: 2025.04.02 GOOGLE LLC
  • EP3895365B1 patent drawingFigure 1
  • EP3895365B1 patent drawingFigure 2
  • EP3895365B1 patent drawingFigure 3

AI summary

Methods, systems, and apparatus, including a method for determining network measurements. In some aspects, a method includes receiving, by a first aggregation server and from each of multiple client devices, encrypted impression data. A second aggregation server receives, from each of at least a portion of the multiple client devices, encrypted conversion data. The first aggregation server and the second aggregation server perform a multi-party computation process to decrypt the encrypted impression data and the encrypted conversion data. Each portion of decrypted impression data and each portion of decrypted conversion data is sent to a respective reporting system.