WO2010131939A1

WO2010131939A1 - Text steganography

Info

Publication number: WO2010131939A1
Application number: PCT/MY2009/000167
Authority: WO
Inventors: Lip Yee Por; Kok Onn Chee; Mei Yin Delina Beh
Original assignee: Universiti Malaya
Priority date: 2009-05-14
Filing date: 2009-10-13
Publication date: 2010-11-18
Also published as: MY147993A

Abstract

A method to hide secret data in cover text is described. Secret data is encoded into base64 string. Next, the cover text length is determined (42) to determine minimum capacity of white spaces required to accommodate secret data. Later, data is converted into bits (44). The bits are hidden into cover text (46) wherein bit value of O is represented with single space between text and bit value of 1 is represented with double space between text to form a stego-text (22).

Description

TEXT STEGANOGRAPHY

The present invention relates generally to secure data processing, more particularly to secure data processing involving information hiding in text files.

BACKGROUND OF THE INVENTION

Hidden message can be relayed from a sender to intended recipient in the form of steganography. Generally a message is hidden in an unsuspected matter such as images, articles, shopping list or cover text. Steganography is different from cryptography. Encrypted messages will arouse suspicion. The hidden messages of steganography do not attract attention.

An early form of steganography involves the application of invisible ink between visible lines of a private letter. Now, steganography includes the concealment of information within computer files.

WO patent application 0025203 teaches a steganographic encoding method for secure transmission or storage of multimedia files. Primary data such as text, image or other digital data is utilized to encode secondary data, such as text, image, or other digital data. Data element from secondary data is performed an operation with the data element of primary data to generate a key element. The primary data may be rearranged according to a predefined or random manner. The key element is then securely transmitted or stored.

US patent application 2007223592 describes a method to hide messages in video files. An N-bit data string is encoded into the video where the string includes bits of having value of 1 or 0. If the first bit of the N-bit string comprises a value of 1 , a change is made to the video to represent a value of 1. If the first bit comprises a value of 0, no change is made to the video to represent a value of 0. The second to Nth bits of N-bit string is also hidden in the video with the same operation.

US patent application 2009003701 describes a method to embed hidden message in a digital image file. The hidden message in a form of text message is encoded into a series of bits and encrypted before it is merged in an image. Less important information of a bit map image is removed and hidden data is injected in its place. The above mentioned solution involves hiding a message having a relatively small file size within a multimedia file having a relatively large file size. Large file size will involve much more processing steps to encode and decode the hidden message. An alternative method is desired to hide message while reducing processing steps.

SUMMARY OF THE INVENTION

The proposed invention describes a method to enable users to communicate by embedding a secret data in an innocent cover text. Secret data can be in the form of text documents, image, audio, video or other digital files.

The cover text can be chosen by a user or auto-generated by the system to hide the secret data. The chosen cover text needs to be verified if it has a size suitable to hide a selected secret data. Otherwise, a cover text shall be generated (making the duplication of the same cover text) to have a size large enough to hide secret data.

White spaces between words and paragraph is used to conceal the secret data. The secret data is first converted into a string by Base64 encoding method. The string is then encrypted to increase the security of secret data. The encrypted data are then transformed into bits. A bit 0 is represented as single space while bit 1 is represented as double space.

A method to decode the secret data is also described so that the intended receiver can view the secret data.

BRIEF DESCRIPTION OF THE DRAWINGS

The invention will now be described in greater detail, by way of an example, with reference to the accompanying drawings, in which:

Fig. 1 illustrates a block diagram of text steganography according to the invention;

Fig. 2A and 2B illustrate a flow chart of encoding a secret data in a cover text;

Fig. 3 illustrates a virtual diagram showing the manipulation of white space in a cover text; and Fig. 4 A and 4B illustrates a flow chart of decoding a secret data from a stego text.

DETAILED DESCRIPTION OF THE DRAWINGS

A proposed text steganography is used by a sender 10 to send a secret data 12 to the intended receiver 14, as shown in Fig. 1. The secret data 12 is embedded in a cover text 21 so that it does not raise any suspicion to the file. The cover text 21 can be any text based file such as text based files with the file extension of ".doc" and ".txt".

The secret data 12 is encoded into Base64 string 16 and go through encryption and Base64 encoding process to become encrypted text 18 with a key before it is embedded into cover text 21 to form a stego text file 22. The stego text file is transmitted into a medium, such as the internet 24. The stego text 22 file can be decoded by the receiver by removing the cover text to unveil the encrypted text 18. The encrypted text is decrypted and decoded by Base64 to become Base64 string 16 and decoded into secret data 12.

The encoding process will be described in detail with reference to Fig. 2A and 2B. First, a sender needs to select a secret data 26 that needs to be sent to the receiver.

After selecting the secret data 12, the sender can opt to define a cover text 27 or the system will auto-generate a cover text 30. The cover text is a typical file with texts.

Next, the secret data is converted into bytes using Base64 encoding 32. Let A be the capacity of the secret data and x be the capacity of the secret data after Base64 encoding, excluding the secret data file extension. The relationship of A and x can be shown in the following relationship.

4A X = 3

The total capacity for the identified secret data is x + α where α is the file extension of the secret data file 34. Should the sender choose to auto-generate a cover text 30, the system will identify if the minimum capacity of white spaces required to accommodate total capacity of secret data, x + α is fulfilled with the following relationship.

y_x ≥ x + a yi is the capacity of white spaces identified from a predefined cover text.

^' The mentioned relationship optimizes the size of generated text file but it may raise suspicion if there are incomplete words or sentence in the cover text. For example a cover text may have the following text:

...the Han Dynasty when incompetent eunuchs deceived the emperor and banished good officials. The government had become extremely corrupt on all levels, leading to widespread deterioration of the empire. During the reign of the penultimate Han emperor, Emperor Ling, the Yellow Turban

In order to alleviate the situation, a minimum capacity of extra text required to construct a full sentence, β is added to the predefined cover text. A new formula is formed as follows:

y_x + β ≥ x + a

Hence, words are added to the cover text to form a complete cover text. An illustrative example is show below to complete the paragraph after text is added to complete the sentence.

... the Han Dynasty when incompetent eunuchs deceived the. emperor and banished good officials. The government had become extremely corrupt on all levels, leading to widespread deterioration of the empire. During the reign of the penultimate Han emperor, Emperor Ling, the Yellow Turban Rebellion broke out under the leadership of Zhang Jiao, who allegedly practiced Taoist wizardry. If a sender defines a cover text 27, instead of auto-generating a cover text, the cover text will be analyzed to identify the capacity of the white space in the cover text. If the defined cover text does not have sufficient white space to accommodate secret data then the cover text will be duplicated to produce sufficient white space. The following duplication formula is used where y₂ refers to the, capacity of white space of cover text and n is the number of duplication required.

ny₂ + O - l) > X + tf ,_n= i, 2, 3...n

Later,_, the sender can choose if the secret data ought to be encrypted 36. If the sender chooses to encrypt the secret data, a user defined public key is used to encrypt the data 38. The encrypted data has a capacity of x + α + 8. Next, the encrypted data is encoded with Base64 encoding 40. The encoded data has a

4(x+a+8) capacity of 3 bytes.

Next, the cover text length is determined 42. For auto-generated cover text, the minimum capacity of white spaces required to accommodate the encoded secret data has the following relationship for the capacity of white space that identified from a predefined cover text, Z₁.

+ 8)

7 s> —

The auto-generated cover text may also be added text to complete the paragraph without arousing suspicion in the following manner.

z + β > ⁴⁽*⁺*^+g)

If the sender defines a cover text, the minimum capacity of the white spaces required to accommodate the encoded data, Z₂ can be identified with the following formula.

After that, the secret data is converted into bit wise dataset 44 before it is hidden into cover text. The bit wise dataset is a stream of bits containing a value of 1 or 0. It is later inserted into the white space in the cover text 46. A value of bit 0 is represented with single space while a value of bit 1 is represented with double space.

To fully make use of white space in a document, tab spaces are added at the end of sentence or between paragraphs which act as invisible words. Bits of 1 or 0 can be hidden between each tab space. Hence, more bits can be embedded at the end of the sentence or between paragraphs. A stego-text is created after the insertion of bits 22.

Two examples of how bits are embedded in cover text are shown in Fig. 3. The first example has a secret data having bit value of 0101 0111 0110 1000 0110 1001 0111.. The first bit 0 is placed between text 'The' and 'government' as single space. The second bit 1 is placed between text 'government' and 'had' as double space. The secret data has 28 bits. The cover text has a capacity of 28 white spaces without tab space.

The second example has a secret data having bit value of 0101 0111 0110 1000 0110 1001 0111 0100 0110 0101 0101 0011. A line separates the first sentence and the second sentence. Tab spaces are fitted between the first sentence and second sentence. There are 20 tab spaces fitted between the lines. The addition of 20 tab spaces allows 48 bits of data to be hidden in the cover text. Using the same cover text, more bits can be inserted because the bits are hidden between tab spaces that used between paragraphs.

Now the decoding process of text steganography will be described. The receiver needs to select the specific stego-text received from sender for decoding 72.

After selecting the stego-text, all the spaces in the text are analyzed to unhide bits 74. Single space between cover text is assigned a bit value of 0 while double space between text is assigned a bit value of 1. The identified bits are combined into a bit wise dataset. The bit wise dataset is later converted into strings of characters 76.

If the string is encrypted, a password is required as input 78 for the string to be decrypted. The strings are decoded with Base64 decoding process 80. The relationship between capacity of strings, B and capacity of bytes data with file extension after Base64 decoding, x is shown below.

^{Λ ~~} 4 The system will then decrypt bytes data with a receiver provided public key 82.

The string data or decrypted data will then be detached into bytes data and the extension of the original secret data 84. Later, the file is decoded with Base64 decoding 86. The secret data 12 is retrieved. '

Accordingly, the invention discloses a method to hide secret data in a cover text. It is the combination of the above features and its technical advantages give rise to the uniqueness of such invention. Modification can be made to assign bit value of 0 as double space and bit value of 1 as single space. If such arrangement were made, then the decoding of secret data will assign a bit value of 0 to double space and a bit value of 1 to single space to unhide bits in a stego-text file. Although the descriptions above contain much specificity, these should not be construed as limiting the scope of the embodiment but as merely providing illustrations of some of the presently preferred embodiments.

Claims

1. A method to hide secret data in cover text comprising the steps of: encoding secret data into base64 string (32); determining cover text length (42) to determine minimum capacity of white spaces required to accommodate secret data; converting data to bits (44); hide bits into cover text (46); wherein bit value of 0 is represented with single space between text and bit value of 1 is represented with double space between text to form a stego-text

(22).

2. A method to hide secret data in cover text according to claim 1 , further comprising: defining a cover text (27); if the defined cover text does not have sufficient white space to accommodate secret data then the cover text will be duplicated, n times to produce sufficient white space; determining cover text length (42) to determine minimum capacity of white spaces required to accommodate secret data from the following formula

wj£ +(«-!) >x+α _{n s=} ^ _{2j 3}..._n where y₂ refers to the capacity of white space of cover text, n is the number of duplication required, x is the capacity of secret data after base64 encoding, excluding secret data file extension, and α is the file extension of the secret data file.

3. A method to hide secret data in cover text according to claim 1 , further comprising: defining a cover text (27); encrypting the secret data (38); if the defined cover text does not have sufficient white space to accommodate secret data then the cover text will be duplicated, n times to produce sufficient white space; determining cover text length (42) to determine minimum capacity of white spaces required to accommodate secret data from the following formula nz₂ Z +( vn-ϊ) / ≥^ψ 3 ^ , n = 1₁,₂2,₃3...n _n where n is the number of duplication, Z₂ is the minimum capacity of white spaces required to accommodate encoded data; x is the capacity of secret data after base64 encoding, excluding secret data file extension, and α is the file extension of the secret data file.

4. A method to hide secret data in cover text according to claim 1 , further comprising: auto-generating cover text (30); identifying if the minimum capacity of white spaces required to accommodate total capacity of secret data is fulfilled; determining cover text length (42) to determine minimum capacity of white spaces required to accommodate secret data from the following formula

y _x > x + a where y-, is the capacity of white spaces identified from a predefined cover text, x is the capacity of secret data after base64 encoding, excluding secret data file extension, and α is the file extension of the secret data file.

5. A method to hide secret data in cover text according to claim 1 , further comprising: auto-generating cover text (30); encrypting the secret data (38); identifying if the minimum capacity of white spaces required to accommodate total capacity of secret data is fulfilled; determining cover text length (42) to determine minimum capacity of white spaces required to accommodate secret data from the following formula

Z ₁ > ^ ⁴ ( ^{x +} ₃ ^{a + 8} )

where Z₁ is the minimum capacity of white spaces required to accommodate encoded data; x is the capacity of secret data after base64 encoding, excluding secret data file extension, and α is the file extension of the secret data file.

6. A method to hide secret data in cover text according to claim 1 , further comprising: auto-generating cover text (30); identifying if the minimum capacity of white spaces required to accommodate total capacity of secret data is fulfilled; adding a minimum capacity of extra text, β which is required to construct a full sentence; determining, cover text length (42) to determine minimum capacity of white spaces required to accommodate secret data from the following formula y_} + β ≥ x + a where y^ is the capacity of white spaces identified from a predefined coyer text, β is the number of added text, x is the capacity of secret data after base64 encoding, excluding secret data file extension, and α is the file extension of the secret data file.

7. A method to hide secret data in cover text according to claim 1 , further comprising: auto-generating cover text (30); encrypting the secret data (38); identifying if the minimum capacity of white spaces required to accommodate total capacity of secret data is fulfilled; adding a minimum capacity of extra text, β which is required to construct a full sentence; determining cover text length (42) to determine minimum capacity of white spaces required to accommodate secret data from the following formula

where Z₁ is the minimum capacity of white spaces required to accommodate encoded data; β is the number of added text, x is the capacity of secret data after base64 encoding, excluding secret data file extension, and α is the file extension of the secret data file.

8. A method to hide secret data in cover text according to claim 1 , further comprising addition of tab spaces at the end of sentences or between paragraphs which act as invisible words.

9. A method to decode secret data in cover text according to claim 1 , comprising the steps of: unhiding bits of stego-file (74) by assigning bit value of 0 to single space and § bit value of 1 to double space; converting bits to string (76); detaching secret data file extension (84); decoding with base64 decoding (86); wherein secret data (12) is retrieved. 0

10. A method to decode secret data in cover text according to claim 9, further comprising decrypting the string (82) if a password is required as input.

11. A method to hide secret data in cover text according to claim 1 , wherein bit value5 of 0 is represented with double space between text and bit value of 1 is represented with single space between text to form a stego-text (22).

12. A method to decode secret data in cover text according to claim 9 and 11 , comprising the steps of unhiding bits of stego-text file (74) by assigning bit value0 of 0 to double space and bit value of 1 to single space.