Blockchain Neural Network Distribution for Security
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Neural networks used in computer security products are vulnerable to adversarial attacks that can poison or evade detection, reducing their accuracy in identifying zero-day threats, and they lack transparency, making it difficult to explain their decision-making processes.
Innovation Solution
The distribution of neural networks using blockchain technology, where each block of the blockchain stores the mathematical functions, model parameters, and data applied to and produced by the neural network, ensuring integrity and allowing only authorized entities to update the network, thereby preventing malicious tampering and providing a transparent history of training and operation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If neural networks are deployed for threat detection, then detection capability for zero-day threats is improved, but vulnerability to adversarial attacks increases
Solution Approach 1:
The neural network model is segmented into multiple components (model structure, model parameters, training data) and stored in separate blockchain blocks. This segmentation allows for granular verification and update control, preventing adversarial attacks by ensuring each component's integrity independently through cryptographic hashing and consensus mechanisms.
Solution Approach 2:
The system performs preliminary actions by storing the trained neural network model and its parameters in the blockchain before deployment. This pre-establishment of the model's cryptographic footprint enables future verification of model integrity and detection of any adversarial tampering attempts.
2Measurement precision
If neural network model parameters are updated to improve detection accuracy, then threat detection performance is improved, but risk of malicious tampering increases
Solution Approach 1:
The system implements feedback mechanisms where the blockchain network verifies model parameter updates through consensus before acceptance. Any attempted malicious tampering is detected through cryptographic verification, providing immediate feedback that prevents adversarial attacks while allowing legitimate accuracy improvements.
Solution Approach 2:
The system applies preliminary anti-action by establishing cryptographic constraints and consensus requirements for model updates before they occur. This pre-established security framework actively prevents adversarial attacks by requiring legitimate authorization and verification for any parameter changes.
3Reliability
If blockchain is used to store neural network data, then security and integrity are improved, but system complexity increases
Solution Approach 1:
The system extracts only the essential cryptographic elements (hashes, consensus markers, authorization data) from the full blockchain protocol and integrates them into the neural network distribution system. This selective extraction maintains data integrity benefits while reducing unnecessary complexity from the overall system architecture.
Data Source
AI summary
Example apparatus disclosed herein generate blocks of a blockchain, the blockchain to store a neural network that has input nodes, hidden nodes and output nodes, with respective ones of the blocks of the blockchain including respective code and respective data to represent corresponding ones of the output nodes of the neural network, a first one of the blocks including first code and first data to implement operations to be performed by a first one of the output nodes, the hidden nodes and the input nodes on input data applied to the neural network to determine an output of the first one of the output nodes. Disclosed example apparatus also train the neural network to determine at least portions of the respective data to include in the respective ones of the blocks of the blockchain, and forward the blockchain to a server that is to distribute the neural network to client(s).