Homomorphic Encryption for Secure Data Reconstruction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data transport methods in distributed networks face computational bottlenecks at target terminals, especially when using redundant codes, and require both decryption and decoding, which can be resource-intensive for mobile devices.
Innovation Solution
Implementing a method that uses encoded data parts from multiple source terminals, where these parts are encrypted using a homomorphic encryption scheme, allowing homomorphic equivalents of arithmetic operations to be performed at a decoder server, reducing the computational load on the target terminal by transmitting results for decryption and decoding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If reconstruction is performed at the target terminal using redundant codes, then data security is maintained, but computational load on the target terminal becomes excessive
Solution Approach 1:
The patent introduces a decoder server as an intermediary between source terminals and target terminal. The decoder server performs the computationally intensive reconstruction operations using homomorphic encryption, allowing the target terminal to retrieve data with minimal computational burden while maintaining security through encrypted operations.
Solution Approach 2:
The patent replaces traditional decryption-and-decoding mechanical process with homomorphic encryption-based computation. This allows arithmetic operations to be performed directly on encrypted data, substituting the need for mechanical decryption steps and reducing computational complexity at the target terminal.
2Reliability
If encryption is applied to data parts at source terminals, then data security is improved, but the target terminal must perform both decryption and decoding which increases computational burden
Solution Approach 1:
The patent replaces the traditional decryption-and-decoding mechanical process with homomorphic encryption-based computation. This allows arithmetic operations to be performed directly on encrypted data, substituting the need for mechanical decryption steps and reducing computational complexity at the target terminal.
Solution Approach 2:
The patent changes the cryptographic parameter state by using homomorphic encryption schemes that allow computations on encrypted data. This transforms the encryption parameter from a state requiring decryption to a state enabling direct computation, thereby improving retrieval efficiency while maintaining security.
3Speed
If distributed uploading from multiple source terminals is used, then network bottleneck is reduced, but joint availability of all source terminals becomes critical
Solution Approach 1:
The patent applies preliminary action by pre-distributing encoded data parts from the data object to multiple source terminals before retrieval is needed. This allows the system to tolerate failures of some source terminals during retrieval, as the necessary data parts are already distributed and encoded to enable reconstruction from subsets.
Solution Approach 2:
The patent changes the data distribution parameter from storing complete data parts to storing encoded data parts through redundant coding. This transformation allows retrieval to proceed with fewer available source terminals, changing the availability parameter from requiring all terminals to requiring only a sufficient subset.
Data Source
Figure 1~2
Figure 3~4
Figure 5~5a
AI summary
A data object is encoded in a redundant code. The redundant code defines a decoding scheme for reconstructing the data object from a sub-set of the encoded data parts. At least the sub-set of the encoded data parts is encrypted using a homomorphic encryption scheme, which allows equivalents of the arithmetic operations of a reconstruction process to be performed on encrypted encoded data parts. The data parts are stored distributed over a plurality of source terminals of a communication network, for use by a target terminal of the communication network. Upon a retrieval command from the target terminal, an upload management module determines which source terminals are available and the upload management module determines causes a selected set of terminals to transmit the encrypted encoded data parts each via its own connection to the network to a decoder server. The decoder server performs homomorphic equivalent operations of arithmetic operations of the reconstruction process and downloads the results to the target terminal. In this way the target terminal does not need to perform the entire reconstruction, without requiring decryption outside the target terminal.