Development of a Hybrid Cryptographic Security Model Combining Lightweight Present and RSA for Resource-Constrained Devices
Abdulkadir Yusuf
Prof. Olanrewaju O.M.
Mr. Mukhtar Abubakar
Abstract
Protecting sensitive data in cloud and edge computing environments remains challenging due to the conflicting requirements of strong security, efficient computation, and support for privacy-preserving processing on resource-constrained devices. Existing lightweight cryptographic schemes provide high efficiency but suffer from weak scalability and key management limitations, while conventional hybrid encryption models improve key security yet rely on decrypt-to-compute paradigms that expose data during cloud-side processing. Conversely, fully homomorphic encryption enables computation over encrypted data but introduces prohibitive computational and memory overhead, limiting its practicality for latency-sensitive edge and IoT systems. To address these limitations, this paper proposes a dual-layer hybrid cryptographic framework that integrates the PRESENT lightweight block cipher with an RSA-based partial homomorphic encryption mechanism. In the proposed architecture, PRESENT is employed for high-throughput bulk data encryption, ensuring low latency and minimal resource consumption, while RSA is used to securely encapsulate the session key and enable controlled encrypted-domain computation through its multiplicative homomorphic property. This design achieves a practical balance between efficiency, security, and functional flexibility without incurring the high cost of fully homomorphic schemes. The framework was implemented in Python and evaluated over 50 independent experimental runs. Performance results demonstrate average encryption and decryption times of 0.031 s and 0.034 s, respectively, with sustained encryption throughput exceeding 6.5 MB/s for typical data payloads. Comparative analysis shows over 40% improvement in computational efficiency relative to existing hybrid baselines while preserving cryptographic correctness and security properties. These results confirm that the proposed framework is well suited for secure, efficient, and privacy-aware cloud and edge computing applications.
References