Scholarlink

<p>In this paper, we propose a lightweight steganography scheme using the Laplacian mathematical model to secure and privacy-preserving communication for edge computing. The contribution proposed in this paper is to meet the challenges of computing overheads associated with the classical cryptographic systems, as well as poor robustness and adaptability for traditional steganographic methods like LSB (Least Significant Bit), DCT (Discrete Cosine Transform), DWT (Discrete Wavelet Transform) and SVD (Singular Value Decomposition). The goal is to build an efficient Laplacian-based embedding model for imperceptible and secure data hiding in low-resource, distributed systems.&#xa0;The Graph-Laplacian operator is utilized to learn the local pixel relations and hide secrets in the high-frequency structural components of the cover image. Lightweight key-controlled embedding scheme provides security and integrity of data, while the modular framework is very flexible to apply over different categories of media such as images, audio, and text. The system is lightweight and efficient for edge devices including Raspberry Pi and Jetson Nano, with low consumption of memory and fast processing.&#xa0;Experiments conducted on standard 512× 512 grayscale benchmark images with a binary “BLOCKCHAIN” watermark show that the proposed method achieves a PSNR of 40.7 dB, SSIM of 0.986, and entropy deviation (ΔH) of 0.017. Compared with LSB-, DCT-, DWT-, SVD-, and hybrid DWT–SVD-based schemes, it delivers improved imperceptibility, efficiency, and robustness (BER ≤ 2.4%, CC ≥ 0.98).&#xa0;The framework provides a practical, secure, and lightweight solution for privacy-preserving data transmission in IoT and decentralized applications.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Lightweight Laplacian Steganography: A Mathematical Framework for Edge-based Privacy Preservation

  • 期刊论文

  • 发表日期：

    2026年1月27日

查看全文
  • R. Gurunath,
  • Debabrata Samanta,
  • Blerta Prevalla Etemi

摘要

In this paper, we propose a lightweight steganography scheme using the Laplacian mathematical model to secure and privacy-preserving communication for edge computing. The contribution proposed in this paper is to meet the challenges of computing overheads associated with the classical cryptographic systems, as well as poor robustness and adaptability for traditional steganographic methods like LSB (Least Significant Bit), DCT (Discrete Cosine Transform), DWT (Discrete Wavelet Transform) and SVD (Singular Value Decomposition). The goal is to build an efficient Laplacian-based embedding model for imperceptible and secure data hiding in low-resource, distributed systems. The Graph-Laplacian operator is utilized to learn the local pixel relations and hide secrets in the high-frequency structural components of the cover image. Lightweight key-controlled embedding scheme provides security and integrity of data, while the modular framework is very flexible to apply over different categories of media such as images, audio, and text. The system is lightweight and efficient for edge devices including Raspberry Pi and Jetson Nano, with low consumption of memory and fast processing. Experiments conducted on standard 512× 512 grayscale benchmark images with a binary “BLOCKCHAIN” watermark show that the proposed method achieves a PSNR of 40.7 dB, SSIM of 0.986, and entropy deviation (ΔH) of 0.017. Compared with LSB-, DCT-, DWT-, SVD-, and hybrid DWT–SVD-based schemes, it delivers improved imperceptibility, efficiency, and robustness (BER ≤ 2.4%, CC ≥ 0.98). The framework provides a practical, secure, and lightweight solution for privacy-preserving data transmission in IoT and decentralized applications.