Encrypted Stream Processing Watermarks for Confidential Event Timing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing stream processing technologies fail to adequately secure confidential data, particularly in distributed environments where administrators can potentially leak sensitive information such as event occurrence times and windows, despite using methods like TEE and homomorphic encryption.
Innovation Solution
An information processing system that performs stream processing on encrypted event information using secure computation to calculate watermarks and windows, employing Trusted Execution Environment (TEE) or homomorphic encryption to protect these sensitive data elements from leakage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If distributed stream processing is performed on encrypted data, then processing capability and scalability are improved, but security of confidential data deteriorates due to potential leakage by administrators
Solution Approach 1:
The patent introduces a watermark as an intermediary mechanism that is embedded in the encrypted data stream. This watermark serves as a mediator between the need for data processing and the need for security, allowing the system to track and detect potential leakage by administrators without requiring them to have access to the actual confidential data content.
Solution Approach 2:
The patent applies the concept of embedding detectable markers (watermarks) within the data stream, analogous to changing color or visual properties. The watermark is embedded in the encrypted data in such a way that it can be detected to identify leakage, providing a security mechanism that does not interfere with the normal processing of encrypted data.
2Reliability
If traditional encryption methods are used in distributed stream processing, then data confidentiality is maintained, but detection of data leakage becomes difficult
Solution Approach 1:
The watermark acts as an intermediary detection mechanism embedded within the encrypted data stream. It enables the system to detect leakage events without requiring decryption or direct inspection of the confidential data content, thus maintaining confidentiality while improving detectability.
Solution Approach 2:
The embedded watermark functions as a detectable marker within the data stream, similar to how color changes can indicate presence or state. This allows the system to detect when data has been leaked or accessed unauthorizedly, transforming an undetectable security issue into a detectable one without compromising data confidentiality.
3Productivity
If administrators have access to process encrypted data streams, then processing efficiency is improved, but risk of insider threats and data leakage increases
Solution Approach 1:
The patent applies preliminary anti-action by embedding watermarks in the encrypted data stream before processing occurs. This preventive measure is in place beforehand to detect and deter potential insider threats, allowing the system to identify and respond to leakage attempts by administrators who might otherwise have unrestricted access to process the data efficiently.
Solution Approach 2:
The watermark serves as an intermediary monitoring mechanism that operates within the data stream processed by administrators. It enables detection of insider threats without requiring administrators to have access to confidential data content, thus maintaining processing efficiency while reducing the harm from potential insider threats.
Data Source
AI summary
An information processing system that performs stream processing on event information that is encrypted and is sequentially input, the information processing system including a watermark calculation unit that calculates a watermark based on first time information included in the event information and second time information obtained by observing or processing the event information by the information processing system, using secure computation.