Time-Adaptive Hybrid Encryption Pipeline For Secure Text Processing Using AES, Homomorphic Encryption, and PQC

Abstract

In cloud environments, secure text processing is implemented using a time-adaptive hybrid encryption approach. Using a traditional encryption model (AES), the loss of control over cloud data creates significant complication; the homomorphic encryption (HE) model incurs processing delays and increased cloud computing costs. The proposed model utilizes a hybrid approach where AES, HE/CKKS, and PQC/Kyber768 are integrated within a cloud environment.

The design and functionality of the hybrid approach are dictated by the Intelligent AI-based router, designed to analyze cloud text processor enabling the hybrid AES model computing cloud’s metadata pertaining to text size, sensitivity, and cloud response time. Secure Enclave environments and HE and AES quantitative partition the data. A tailored functionality partitions the HE and AES environments. A blockchain-based audit trail encrypts and secures all primary events, thus providing a permanent record of all integral cryptarithmic transfers.

The model in question performed processing of a representative text workload of size 0.39MB. The results indicated a total server-side processing duration of 347.40 ms, coupled with a client-side decryption duration of 14.60 MS, leading to a secure cycle duration of below 0.4 seconds. The level of encrypted text size expansion was limited to 1.36× (previous models exhibited a much larger expansion), with a transfer rate of 1.46MB/s. The results of homomorphic encryption exhibited 100 percent accuracy, which confirms that the model successfully balanced performance, storage efficiency, and optimal security of approximately 192 bits, with flexibility concerning potential future quantum threats.