A DNA Based Lightweight Cryptography Framework Integrating ECC and RNN Based Deep Learning for IoT Security
The exponential growth of Internet-of-Things (IoT) networks and connected devices has highlighted critical challenges in ensuring high security, low latency, and low energy consumption for sensitive data exchange in resource-constrained environments. Traditional cryptographic systems often struggle...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | en en en |
| Published: |
University of Malaysia Sarawak
2025
|
| Subjects: | |
| Online Access: | http://ir.unimas.my/id/eprint/49935/1/dsva_Sehrish%20Aqeel.pdf http://ir.unimas.my/id/eprint/49935/2/Thesis%20PhD_Sehrish%20Aqeel.pdf http://ir.unimas.my/id/eprint/49935/3/Thesis%20PhD_Sehrish%20Aqeel_24%20pages.pdf http://ir.unimas.my/id/eprint/49935/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The exponential growth of Internet-of-Things (IoT) networks and connected devices has highlighted critical challenges in ensuring high security, low latency, and low energy consumption for sensitive data exchange in resource-constrained environments. Traditional cryptographic systems often struggle to meet these demands, necessitating innovative solutions tailored to IoT. This study proposes a novel DNA-Based Lightweight Cryptographic System (DNA-LWCS) integrating the inherent randomness of DNA sequences with Elliptic Curve Cryptography (ECC) to enhance security and efficiency. Leveraging the robustness of DNA sequences and ECC’s computational efficiency, DNA-LWCS provides robust protection against attacks like Distributed Denial-of-Service (DDoS) and Man-in-the-Middle (MITM), addressing IoT vulnerabilities. Deep learning models, including Recurrent Neural Networks (RNN) and Gated Recurrent Units (GRU), achieve remarkable precision of 99.62% for RNN and 98.61% for GRU in key strength estimation, optimizing encryption processes. Experimental results demonstrate DNA-LWCS outperforms 3DES, AES, ECC, and benchmarks in critical security and efficiency metrics. DNA-LWCS achieves a correlation coefficient of 0.025, ensuring high confidentiality and data integrity, fast encryption times (e.g., 0.800 seconds for Photographer), high throughput (12.00 Bps at 50 data points), and low energy consumption (10.00–15.00 J), ideal for resource-constrained IoT devices like smart meters and smart homes. DNA-LWCS demonstrates superior entropy values (7.800 for Lena, 6.500 for Photographer) and avalanche effect (52.99% for Lena, 50.99% for Photographer), meeting stringent randomness and input sensitivity criteria for applications like autonomous vehicles and patient monitoring. This research advances IoT security by combining DNA-based encryption with deep learning, delivering a scalable, efficient, and resilient cryptographic framework that bolsters data protection and fulfills the essential requirement for lightweight, energy-efficient encryption in contemporary interconnected systems. |
|---|