Coverless Information Hiding: A Web Text-Based Covert Communication System with Multi-Feature Encoding and Layered Security

Authors

  • S. Suneel Kumar Author
  • G. Nagireddy Author
  • B. Revanth Author
  • S. Prem Chandhu Author
  • M. Gowtham Author

DOI:

https://doi.org/10.62643/ijerst.2026.v22.n2.pp128-135

Keywords:

Coverless Information Hiding, Steganography, Web Text, Covert Communication, Steganalysis Resistance, Natural Language Processing, Data Security, Flask Web Application, Multi-Feature Encoding

Abstract

Digital communication security has historically oscillated between cryptographic confidentiality and steganographic undetectability. Both paradigms share a critical weakness: cryptography reveals that a secret exists, while conventional steganography modifies a carrier medium, leaving statistical fingerprints that modern steganalysis tools reliably detect. This paper presents a coverless information hiding system that transcends both limitations by transmitting secret messages entirely through the selection and arrangement of unmodified web text retrieved in real time from the live internet. No carrier file is ever altered, rendering conventional statistical detection theoretically inapplicable. The proposed system encodes a message by mapping each character’s ASCII ordinal to a sentence-feature target value, generating a contextual search query, scraping live web pages via BeautifulSoup4, and selecting the sentence whose multidimensional feature vector—comprising word count, character count, vowel density, and punctuation frequency— most closely satisfies the encoding constraint. Decoding reverses this mapping deterministically without requiring a pre-shared corpus, eliminating the corpus-synchronisation vulnerability that affects prior art. A Flask-based web application with bcrypt password authentication, Flask-Login session management, AES-based optional encryption, and SQLite persistence provides a complete deployment platform. Empirical evaluation across 200 test messages demonstrates 96–98% feature-extraction accuracy, encoding throughput of 3–7 bits per sentence, and end-to-end decoding latency of under 0.5 seconds. Comparative analysis against traditional LSB steganography (StegHide, OpenStego) and corpus-based coverless systems confirms superior undetectability with competitive capacity. All 45 test cases across unit, integration, system, performance, and security dimensions passed, validating the framework for secure, undetectable covert communication in government, corporate, journalistic, and personal-privacy contexts.

Downloads

Published

02-04-2026

How to Cite

Coverless Information Hiding: A Web Text-Based Covert Communication System with Multi-Feature Encoding and Layered Security. (2026). International Journal of Engineering Research and Science & Technology, 22(2), 128-135. https://doi.org/10.62643/ijerst.2026.v22.n2.pp128-135