A Novel Approach to Cryptography: Deep Learning-based Homomorphic Secure Searchable Encryption for Keyword Searches in the Blockchain Healthcare System

Security is the most important element of encryption over the Internet because of the value and importance of patient health records (PHR). Users that use keyword searches to obtain access to the database's PHR are more vulnerable to security issues. Despite the fact that a blockchain-based healthcare system can ensure security, current schemes have significant limitations. Existing methods have solely focused on data storage, with blockchain serving as a storage database. In this study, we created a secure searchable blockchain based on deep learning as a distributed database that uses homomorphic encryption to allow users to securely access data via search. Secure key revocation and update rules will be progressively included in our proposed research. This study uses an IoT dataset to analyze and compare our proposed access control mechanisms to benchmark models. The proposed techniques are implemented in the Hyperledger tool utilizing smart contracts. The proposed strategy is compared to others already in use. In comparison to benchmark models, our proposed solution considerably increases security, anonymity, and user behavior tracking, resulting in a more efficient blockchain-based IoT system.