Literature Review & Prior Art Matrix

A rigorous academic review of 16 prior text steganography techniques was conducted to identify structural bottlenecks, capacity limitations, and security vulnerabilities.

1. Analysis of the 16 Prior Approaches

1. Emoji Encoder (Paul Butler, 2025)

Appended Variation Selectors behind emojis (1 byte/VS). While offering good capacity and clipboard robustness, it lacked compression, lacked encryption, caused emoji rendering glitches across platforms, and was trivial to detect using simple regular expressions.

2. Zero Width Characters (Priyansh Sharma, 2023)

Used ZWC (\u200C, \u202C, \u200E, \u202D) mapped to 2-bit blocks with fixed XOR keys. Resulted in low capacity, weak security, and language distortion in right-to-left scripts.

3. Multilayer FPE + Huffman (Majid et al., 2022)

Divided payload into 3-bit ZWC blocks with Format-Preserving Encryption. Achieved 18.4% capacity ratio, but severely distorted Arabic and RTL text due to Left-To-Right Embedding (LRE) control characters.

4. AITSteg (Ahvanooey et al., 2018)

Employed Gödel numbering and time-derived XOR keys. Sensitive to transmission delays, required 6 ZWC characters per single secret character, and had weak cross-platform stability.

5. SNOW Tool (Matthew Kwan, 2013)

Appended whitespace and tabs at line ends with ICE encryption (3 bits/group). Highly vulnerable to social media sanitizers that strip trailing spaces and tabs automatically.

6. Binary Bit Mapping OTP (Naharuddin et al., 2018)

Matched secret bits with cover ASCII bits using One-Time Pads. Achieved 14.2% capacity but generated massive stego-keys that scaled with secret length, and failed if a single cover character was edited.

7. Bulk Data Hiding via Chaotic Sequences (Shivani et al., 2015)

Used word abbreviations and chaotic matrix matching. Required enormous cover text sizes to satisfy bit matching probability.

8. ASCII Art Steganography (Vincent Chu, 2009)

Converted images to text art. High capacity (~40KB) but visually bizarre and instantly flagged by observers.

9. 3-Bit Covert Channel Format (Gurunath & Samanta, 2023)

Modified Word/HTML font sizes (0.5pt) and vertical positions. Lost immediately upon copying to plain text.

10. Character Spacing Normalization (Syed Tahir Ali Shah et al., 2020)

Modulated letter spacing by 0.1–0.4pt. Fragile against plain text copy, English-only, and destroyed in print/chat.

11. PDF Justification Steganography (Khosravi et al., 2019)

Manipulated space ordering in justified PDFs. Very low capacity (1/246 ratio) and dependent on OCR accuracy.

12. Email LZW Color Coding (Malik et al., 2016)

Colored email header and body text. Restricted strictly to HTML email clients.

13. Word Font Type and Style (Kumar et al., 2016)

Mapped 7 bits per space across 64 fonts in MS Word. Metadata lost on plain text clipboard copy.

14. Format-Based Shifting (Roy & Manasmita, 2011)

Line and word shifting with $O(n^2)$ computational complexity and reliance on OCR scans.

15. Emoticon-Based Steganography (Ahmad et al., 2014)

Mapped secret strings to 256 emoji dictionaries. Created unnatural emoji spam in low-frequency text.

16. Email Repeating Characters (Fateh & Rezvani, 2018)

Generated recipient lists based on text frequency. Network anomaly due to excessive synthetic email addresses.

2. Comparative Analysis Matrix (Table 2.11)

3. Identified Research Gap Resolved by STEGNOLINES

Prior literature lacked an integrated system balancing high capacity, cryptographic security, and multi-platform survival. STEGNOLINES bridges these gaps through:

• Brotli WASM Compression: Overcomes the lack of compression in Butler and Sharma.
• AES-256-CTR & PBKDF2: Replaces weak XOR and legacy ciphers with NIST-standardized cryptography.
• Pure Variation Selectors: Eliminates the RTL and Arabic distortion introduced by ZWC markers.