Reinforcement Learning Encryption for Self-Destructing Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Self-destructing schemes in cloud computing environments face challenges in ensuring data is deleted at the desired time, leading to potential security and availability issues.
Innovation Solution
A reinforcement learning-based encryption and decryption method that manages encryption keys and threshold values to optimize data availability and security, using a client and server system with modules for key management, secret sharing, and threshold estimation to ensure data is deleted automatically within the desired timeframe.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If a self-destructing scheme is used to protect user data in cloud computing, then privacy protection is improved, but data availability may deteriorate because data may be deleted before the user desires
Solution Approach 1:
The patent applies reinforcement learning to dynamically adjust the threshold value parameter, which determines when data should be deleted. By learning from environmental feedback and optimizing the threshold based on reward functions, the system adapts the deletion timing parameter to balance privacy protection with data availability, preventing premature deletion while ensuring eventual data destruction.
Solution Approach 2:
The patent implements a feedback mechanism through reinforcement learning where the system observes the outcomes of deletion decisions and adjusts its policy accordingly. The reward function provides feedback on whether data was deleted at the appropriate time, allowing the system to learn from past decisions and improve future threshold setting to maintain both privacy and availability.
2Reliability
If the threshold value for data deletion is set low to ensure data availability, then data can be accessed longer, but security deteriorates because data may not be deleted in time
Solution Approach 1:
The reinforcement learning system dynamically adjusts the threshold value parameter based on learned patterns and environmental feedback. Instead of using a fixed low threshold that compromises security, the system learns to set optimal thresholds that balance availability and security requirements, adjusting the parameter adaptively rather than statically.
Solution Approach 2:
The reward function provides feedback on security outcomes, allowing the system to learn from instances where data should have been deleted but wasn't. This feedback mechanism enables the system to improve its threshold setting over time, preventing security violations while maintaining necessary data availability.
3Reliability
If reinforcement learning is applied to estimate the threshold value, then data lifetime and availability are optimized, but system complexity increases
Solution Approach 1:
The reinforcement learning system is self-adjusting and learns autonomously from environmental feedback without requiring manual configuration or intervention. The threshold value is automatically optimized through the learning agent's interaction with the environment, reducing the need for complex manual tuning while achieving optimal data lifetime and availability.
Solution Approach 2:
The system performs preliminary learning during an exploration phase where the reinforcement learning agent tries different threshold values and learns from the outcomes. This preliminary action allows the system to build knowledge before actual data deletion decisions are made, reducing complexity during operational phase by relying on pre-learned policies.
Data Source
AI summary
A client and server system that performs a reinforcement learning-based encryption and decryption method according to an aspect of the present invention may include: a key management module configured to manage an encryption key required in performing an encryption and a decryption of data; a secret sharing module configured to perform a secret sharing of a threshold value for a lifetime and availability of the data; and a threshold estimation module configured to perform an estimation of the threshold value; and can improve the availability and security of data to satisfy user demands in a self-destructing environment for privacy protection.