Encrypted data repair based on Reed-Solomon and machine learning

In this paper, a two-layer protection system RS-ML combining Reed-Solomon (RS) error correction code and machine learning (ML) is implemented and evaluated, and the data recovery effect of the system under different damage rate scenarios is studied to ensure data security and integrity. Through experiments, this paper analyzes the recovery rate at a damage rate of 1% to 20%. The experimental results show that machine learning enhancement provides an additional recovery layer for damaged data. At low damage rates (1%, 5%), the overall recovery rate is maintained at more than 99%. Under the medium damage rate (10%, 12%), it can still maintain a good recovery rate. This design provides an idea for robust data storage and transmission systems in harsh environments.