A Lightweight DNA-Chaos Hybrid Encryption Framework for Real-Time IoT Image Security

Abstract

The rapid proliferation of Internet of Things (IoT) devices—projected to exceed 40 billion by 2030 and generate over 175 zettabytes of data—makes the need for encryption solutions that are both secure and resource-efficient more critical than ever. Traditional cryptographic algorithms such as AES, DES, and RSA, while highly secure, are often too computationally heavy for low-power IoT environments. In this study, we present a lightweight and high-security image encryption framework that integrates DNA computing with two-dimensional chaotic maps, specifically tailored for real-time IoT applications. The proposed method dynamically encodes image pixels into DNA sequences, generates chaotic key streams, and performs nucleotide-level operations to achieve high randomness, strong diffusion, and confusion. The framework, implemented in Python 3.12 using NumPy and Pillow, was evaluated on 256 × 256 RGB images. It achieved encryption and decryption times of 68 ms and 65 ms, well within real-time processing limits, and delivered a Peak Signal-to-Noise Ratio (PSNR) of 46.82 dB and a Structural Similarity Index (SSIM) of 0.9874, indicating near-perfect image reconstruction. Statistical and differential attack resistance was confirmed through an NPCR of 99.61%, UACI of 33.75%, and histogram entropy of 7.98 bits/pixel. With a low computational footprint—consuming only 50 MB of RAM and 20% CPU - the proposed framework is ideal for edge-based IoT deployments, including smart surveillance, wireless sensor networks, and medical image transmission. This work provides a balanced solution to the dual challenges of robust security and lightweight performance, making it a promising candidate for next-generation secure IoT communications