Encrypted Speech Search Using Phonetic Encoding and Trapdoor
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current cryptographic methods for secure data processing in cloud computing, especially for sensitive data like speech, face challenges in maintaining privacy as they require decryption for processing, which compromises security and is computationally demanding, especially for big data tasks and speech recognition.
Innovation Solution
A method and system that uses phonetic encoding and searchable symmetric encryption to enable encrypted search of speech data, separating acoustic and language model training between client and server sides, and employing AES encryption, along with a Convolutional Neural Network for phonetic symbol classification, to maintain privacy by processing encrypted data without decryption on the cloud.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If data is decrypted for processing in the cloud, then computational efficiency is improved, but data security and privacy are compromised
Solution Approach 1:
The patent introduces an intermediary encryption scheme where the cloud provider possesses a trapdoor that enables efficient processing of encrypted data without full decryption. The trapdoor acts as a mediator that allows the cloud to compute on encrypted speech data while the client maintains control through secret keys, thus achieving both computational efficiency and data security.
Solution Approach 2:
The encryption system is segmented into multiple components: public keys for encryption, secret keys for decryption, and trapdoors for efficient processing. This segmentation allows different parties (client and cloud provider) to have different levels of access, enabling the cloud to process data efficiently without compromising overall security.
2Reliability
If strong encryption is used to protect sensitive data, then data security is improved, but processing speed and computational efficiency deteriorate
Solution Approach 1:
The trapdoor mechanism serves as an intermediary that bypasses the computational burden of full decryption while still enabling efficient processing. The cloud provider uses the trapdoor to perform operations on encrypted data at speeds comparable to processing unencrypted data, thus maintaining both strong encryption and high processing speed.
3Productivity
If cloud providers can process encrypted data efficiently, then productivity is improved, but the risk of unauthorized access and privacy breaches increases
Solution Approach 1:
The system employs asymmetric cryptographic relationships where the cloud provider has a trapdoor for efficient processing but lacks the secret key for full decryption. This asymmetry ensures that even if the cloud provider attempts unauthorized access, they cannot fully compromise the encrypted data, thus enabling efficient processing while mitigating unauthorized access risks.
Solution Approach 2:
The client provides feedback control by holding the secret key and being able to decrypt and verify the data. This feedback mechanism ensures that the cloud provider's processing activities remain under client supervision, allowing efficient processing while maintaining security through client verification.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A system and method for preserving the privacy of data while processing of the data in a cloud. The system comprises a computer program application and a client encryption key, The system is operable to encrypt the computer program application and data using the client encryption key; upload the encrypted computer program application and encrypted data in the cloud; enable the computer platform to undertake processing of the encrypted data in the cloud using the encrypted computer program application; output encrypted processing results; and, enable decryption of the encrypted processing results using the client encryption key.