New approach for text based steganography

(1)

1

T.C.

SELÇUK ÜNİVERSİTESİ

FEN BİLİMLERİ ENSTİTÜSÜ

NEW APPROACH FOR TEXT BASED STEGANOGRAPHY

Abdikarim Abi HASSAN YÜKSEK LİSANS TEZI

Bilgisayar Mühendisliği Anabilim Dalını

(2)

(3)

ii

TEZ BİLDİRİMİ

Bu Tezdeki bütün bilgilerin etik davranış ve akademik kurallar çerçevesinde elde edildiğini ve tez yazım kurallarına uygun olarak hazırlanan bu çalışmada bana ait olmayan her türlü ifade ve bilginin kaynağına eksiksiz atıf yapıldığını bildiririm.

DECLARATION PAGE

I hereby declare that all information in this document has been obtained and presented in accordance with academic rules and ethical conduct. I also declare that, as required by these rules and conduct, I have fully cited and referenced all material and results that are not original to this work.

İmza

Abdikarim Abi HASSAN Tarih: 08.06.2015

(4)

iii

ÖZET

YÜKSEK LİSANS

NEW APPROACH FOR TEXT BASED STEGANOGRAPHY

Abdikarim Abi HASSAN

Selçuk Üniversitesi Fen Bilimleri Enstitüsü Bilgisayar Mühendisliği Anabilim Dalı Danışman: Yrd.Doç.Dr.Kemal TÜTÜNCÜ

2015, 79 Sayfa Jüri

Danışman: Yrd.Doç.Dr.Kemal TÜTÜNCÜ

Yrd.Doç.Dr.Bariş KOÇER Yrd. Doç. Dr. Abdullah TÜMER

Bilgi güvenliği ve özellikle gizli iletişimler gizli mesajlar için bir kaç metodun geliştirilmesine yol açmıştır. Bu metotlar arasında modern steganografi yeniler arasında yer almaktadır. Steganografi bilgi saklama sanatı olarak bilinir. Tarihi M.Ö. 440 yıllarına dayanmakla birlikte binlerce yıldır kullanılmaktadır.

Modern steganografi metin, resim, ses ve video dosyalarına uygulanabilmektedir. Bununlar birlikte metin steganografisi metin dosyalarında anlama katkısı olmayan artık bit olarak tabir edilen kısımları bulmanın zorluğu ve diğer ortamlara göre daha az bilgi saklama kapasitesine sahip olmalarından ötürü son yıllarda daha az ilgi çekmektedir. Bunun ise dikkate alınmaması gerekmektedir. Nedeni ise metin steganografisinin diğer ortamlara göre daha avantajlı olduğu noktaların olmasından ötürü etkin steganografi modeli olarak öne çıkmasıdır.

Metin steganografisinin resim ve ses steganografisine göre bir avantajı bu iki ortamın artık bilgileri kullanma yöntemlerinden ötürü sıkıştırılması ve/veya arık bilgilere bilgi gizlenmesinden ötürü dikkat çekmesidir. Bu durum metin steganografisinde geçerli değildir. Metinlerde artık bilgi olmasına rağmen yok edilemez ya da sıkıştırılamazlar. Ayrıca metin steganografisinde kayıpsız sıkıştırma teknikleri uygulanır. Buda beraberinde metinde kayıpların olmamasına ve daha az dikkat çekilmesine imkan sağlar. Meting steganografiyi resim ya da ses steganografisi yerine seçmeyi sağlayan bir diğer avantaj ise daha az yer kaplamaları ve daha basit iletişim imkanı sağlamalarıdır. Metin tüm dünyadaki iletişimde gerek dijital gerekse de basılı formda en çok kullanılan ana formlardan birisidir. Bu açıdan metinlere ulaşamayan insan sayısı çok fazla değildir.

Bu çalışmada kayıpsız sıkıştırma tekniklerinin kombinasyonu ve Vigenere şifresi e-posta tabanlı metin steganografisi oluşturulmasında kullanılmıştır. Gizli mesajın E-posta gövdesine (örtü metni) gömülmesi/çıkarılması için e-posta adresleri anahtar olarak kullanılmaktadır. Gizli mesaj için en fazla tekrarlı örnek sağlayan örtü metni seçildikten sonar uzaklık matrisi oluşturulur. Uzaklık matrisinin elemanları kayıpsız sıkıştırma algoritmaları olan ve yazıldığı sırada birbiri ardına kullanılan Çalışma Zamanı Kodlama (RLE) + Burrows Tekerlek Dönüşümü (BWT) + Öne Doğru Kaydırma (MTF) + Çalışma Zamanı Kodlama (RLE) + Aritmetik Kodlama (ARI) algoritmalarıyla sıkıştırılmıştır. Sonrasında Latin karesi stego anahtar 1’i ve Vigenere şifresi stego anahtar 1’I bulmayı zorlaştırmak için kullanılmıştır. En son adımda stego anahtar 1ve stego anahtar 2 gizli mesajı e-posta platformuna gömmek için kullanılacak olan e-posta adresilerinin seçimi için kullanılmıştır. Deneysel sonuçlar önerilen metodun makul gizleme kapasitesine ve yüksek güvenlik/karmaşıklık değerine sahip olduğunu göstermiştir.

Anahtar Kelimeler: Metin steganografisi, Latin karesi, Vigenere şifreleyicisi, Stego anahtar, BWT, MTF, RLE, ARI

(5)

iv

ABSTRACT

MS THESIS

NEW APPROACH FOR TEXT BASED STEGANOGRAPHY

Abdikarim Abi HASSAN

THE GRADUATE SCHOOL OF NATURAL AND APPLIED SCIENCE OF SELÇUK UNIVERSITY

THE DEGREE OF MASTER OF SCIENCE / DOCTOR OF PHILOSOPHY IN MECHANICAL ENGINEERING Advisor: Yrd.Doç.Dr.Kemal TÜTÜNCÜ 2015, 79 Pages Jury Advisor Yrd.Doç.Dr.Kemal TÜTÜNCÜ Yrd.Doç.Dr.Bariş KOÇER Yrd. Doç. Dr. Abdullah TÜMER

The case of safety of information and especially secret communications has led to the introduction of several methods for secret message. Among these methods modern steganography is a rather new method. Steganography, the art of information hiding, has been used around for thousands of years, with the earliest examples coming from as early as 440 B.C.

Modern steganography can be applied to text, images, audio and video. However, text steganography, has received less interest recent years, primarily due to the difficulty in finding redundant bits in text files and the lower capacity to hide information than the other mediums. This should not be the issue, as text steganography has many advantages over the other mediums which makes it model for effective steganography.

One advantage of text steganography over image and audio is that while they are both susceptible to compression and/or hiding information due to their use of redundant data, this is not an issue with text steganography as even though text contains redundancy, it cannot be removed or compressed. Additionally only lossless compression is used for text and this let text have no loss and less suspicious. The other advantage to prefer text steganography over images and audio is its smaller memory occupation and simpler communication. Text is also still one of the major forms of communication in the world, both in digital and printed form, and there are not many people who do not have access to text.

In this study combination of lossless compression techniques and Vigenere cipher was used in the e-mail based text steganography. It makes use of email addresses to be the keys to embed/to extract the secret message into/from the email text (cover text). After selecting the cover text that has highest repetition pattern regarding to the secret message the distance matrix was formed. The members of distance matrix were compressed by following lossless compression algorithms as in written sequence; Run Length Encoding (RLE) + Burrows Wheeler Transform (BWT) + Move to Front (MTF) + Run Length Encoding (RLE) + Arithmetic Encoding (ARI). Later on Latin Square was used to form stego key1and then Vigenere cipher was used to increase complexity of extracting stego key1. Final step was to choose mail addresses by using stego key1 and stego key2 to embed secret message into forward e-mail platform. The experimental results showed that proposed method has reasonable performance in terms of capacity and also higher security in terms of complexity.

(6)

v

ACKNOWLEDGEMENT

Before I get into thick of things, I would like to add a few heart-felt words for people who were a part of this thesis in numerous ways, people who gave their unending support from starting stages to the completion of this project report. First acknowledgements are for my thesis Guide Assist.Prof.Dr. Kemal TÜTÜNCÜ, who gave me the opportunity to take up the thesis on the topic “New Approach for Text Based Steganography”, as part of my study environment.

I would also like to thank my parents, who were very helpful and understanding in many a situations. Their support and encouragement helped me a lot in completing the project successfully.

Last but not the least, I would like to also acknowledge all my friends and well wishers, who helped me and encouraged me in some or the other way, by giving their views on the work done by me.

Abdikarim Abi Hassan

(7)

vi

The most significant thing that helped human’s development in all the parts of life and works is considered communication procedure that accomplishes approaches to facilitate the process of communication among different peoples. And one of the oldest communication ways that peoples used as sending a message for an event or accident was durum, smoke, and also drawing coves. Then the humans idea expanded by creating way that allow of writing by utilizing specific symbols named as letters, when was invented the letters then, the telegraph was the second step to the way of growth, and facilitates the communications process afterword, the radio, and television was invented and finally internet was the latest and modern innovations of communication processes.

After the development of new technologies and the expansion of computer systems and the growth of its use in different areas of life and work, also the invention of internet and expansion of number of users in the world, since the internet is a public network the problem of security of information has gained special significance. For this reason, various methods including steganography, cryptography, coding and so on, have been proposed to use.

Steganography is the art and science that hides data inside other data in such a way that no one apart from the intended recipient knows of the existence of the message. This is the main difference between steganography and other methods of hidden exchange of information. However the information to be hidden is called the secret message and the medium in which the information is hidden is called the cover object. The cover object containing hidden message is called stego-object. The algorithms employed for hiding the message in the cover object at the sender end and extracting the hidden message from the stego-object at the receiver end is called stegano system.

Steganography is one of the methods which have attracted more consideration during the recent century. Most steganography tasks have been performed on images, video clips, sound and text. But text steganography is the most difficult kind of steganography; this is due to the very less of redundant information in a text file, while there is a lot of redundancy in a picture or a sound file.

(11)

2 Since there are a great number of message sender or receiver in the Internet and conducting a lot of communications in them, stenography in sentences exchanged in these communications is an appropriate choice for establishing hidden communications.

1.1 History of Steganography

The word steganography is derived from Greek word which means “cover writing”. Most historical stories about stenography are traced its original as far back as 440 B.C., where Herodotus wrote in his histories about two examples of steganography. The first one is Histious shaved the head of one of his most trusted slaves or prisoners and then wrote secret messages on his scalp. Once the hair had reground, the message was hidden and the prisoner could be sent to deliver the message to the other side where no one could read that message (Shirali-Shahreza S. and Shirali-Shahreza M., 2008c).

The second is Demaratus a Greek in the Persian court who sent warning of a forth coming invasion by Xeroxes by writing massage on a wooden pallet and then covering it in wax. The cover tablet was sent to the receiver who would remove the wax and read the hidden message, on the other hand in 20th century; invisible inks were generally used technique. In the Second World War people used natural items to write secret messages for example milk, vinegar, fruit juices and urine (Singh M., Singh S., Singh B. and Devi S., 2007).

During and after World War II, espionage agents used photographically produced microdots to send information back and forth. Microdots were typically minute, approximately less than the size of the period produced by a typewriter. World War II microdots needed to be embedded in the paper and covered with an adhesive, such as collodion. This was reflective and thus detectable by viewing against glancing light. Alternative techniques included inserting microdots into slits cut into the edge of post cards (Rahate N. D. and Rothe P. R., 2014)

During WWII, Velvalee Dickinson, a spy for Japan in New York City, sent information to accommodation addresses in neutral South America. She was a dealer in dolls, and her letters discussed the quantity and type of doll to ship. The stegotext was the doll orders, while the concealed "plaintext" was itself encoded and gave information about ship movements, etc. Her case became somewhat famous and she became known as the Doll Woman.

(12)

3 Jeremiah Denton repeatedly blinked his eyes in Morse Code during the 1966 televised press conference that he was forced into as an American POW by his North Vietnamese captors, spelling out the word, "T-O-R-T-U-R-E". This confirmed for the first time to the U.S. Military (naval intelligence) and Americans that American POWs were being tortured in North Vietnam

(Wikipedia, 2015).

Cold War counter-propaganda. In 1968, crew members of the USS Pueblo intelligence ship held as prisoners by North Korea, communicated in sign language during staged photo opportunities, informing the United States they were not defectors, but were captives of the North Koreans. In other photos presented to the US, crew members gave "the finger" to the unsuspecting North Koreans, in an attempt to discredit photos that showed them smiling and comfortable.

1.2 Motivation

Information security especially steganography face on less consideration from the researchers in recent years while cryptography system has been spent more time. As will be talk in later, this should not be case as steganography can be completely strong tool, particularly if utilized in hybrid with cryptography. However, text steganography is the least popular type of steganography according to the profitable products and researchers, so, there are various conceivable purposes behind this, including a lower capacity, this means text has low redundancy and not possible to embed more text, so that the capacity of text steganography is very low while others like images and audio are embeds more capacity of secret. But surely, there are several reasons why it should again more considerations. One of the reasons that due again for consideration of text steganography are today’s world still using text as a procedure of communication including printed papers, simple messaging service on mobiles, and e-text like email. Also the opportunity to acquire computers and access to the internet may be very limited, for this trouble, a steganography procedure must be appropriate to be making by hand if required. Today, most universally applicable medium for communication is a text. Anyone with access to any digital images or audio of any kind also has access to some structure of a text. This can be digital or printed text and it can be in the forms of emails or sms messages. Lastly, text steganography, particularly the semantic form neglecting in tensed structure and meaning can

(13)

4 withstand outside influences or process than other mediums. encouragement behind developing text steganography approaches according to its utilization in different associations to communicate between its members, and in addition, it can be utilized for communication between members of the national military or intelligence workers or for private companies to conceal secret information or in the area of spying. The key objectives of utilizing steganography are to keep away from adversary attention for the transmission of hidden message.

1.3 Aim

This study will illustrate or focus a particular kind of issue which is labeled or classified under the text steganography problem, thus far, here are no ideal system used to tackle this issue. However, numerous offered or proposed algorithms present us great results. It will concentrate on concealing the secret message in the natural texts.

The most significance parts of Steganography system are the steganographics capacity and stego texs impercebility. Thus, here will be explained basic building block of information hiding which contend part of them: capacity refers to the amount of data bits that can be embedded by cover medium. Security is concerned with the ability to extract the hided information by an observer. Robustness related to resisting ability against destroying or modifying the concealed information.

There are two requirements to conceal data with considering the characteristics of textual documents: First, the readability of textual information should not be affected by the process of hiding. Secondly, according to the visual side the content should not be included within any abnormality.

(14)

5

Figure 1.1: The block of proposed method

In this study, to solve capacity and security problems, new approach of text steganography based on hybrid compression are proposed. Thus, the purpose of this study is to increase the capacity of hidden data (cover medium) while complicating the extracting process of concealed data. For this purpose the message is hidden inside a text selected from prebuilt text base. This text-base, includes: texts that can be used in group messages (speeches) such as: remainder messages, abstracts, notes.

P la in te x t Cover Text Selecting Textual Base Compression via RLE+BWT+MT F+RLE+ARC Latin Square E-mail Addresses Stego-text Vigener Cipher Combining the cover text with E-mail addresses

(15)

6 During the embedding process, the originality of selected cover medium is protected by only camouflaging the secret information. E-mail has been selected as a communication channel, because, communication happens between the two sides.

Therefore, stego medium is arranged in a form of forward mail. By benefiting from email address lists which have already been created. The block of the proposed system can be seeing in Fig.1.1.

1.4 Cryptography

Cryptography is the combination of two Greek words crypto which means “secret” and Graphy which means “writing”. So, cryptography is a way to changing the message/information from one form to another secret form which is differ from the original with the help of a secret key and this process is called Encryption ( Sethi N. and Sharma D., 2012). However, we use the term cryptography to refer to the science and art of transforming messages to make them secure and immune to attacks. The changed value of secret message is called cipher and to get original message from cipher is called decryption.

1.4.1 History of cryptography

About 4000 years ago, cryptography was used by Egyptians on the tomb of their master Nobleman Khunmhotep to write his life story on his tomb. They used symbol instead of words/characters and technique was called substitution method (Sethi, N. and Sharma D., 2012), In 500 BC, a machine named Scytale was developed by Spartans. This machine uses a cylinder to hide the secret message. The message was written in the encrypted form on a tape in such a manner so that if this tape is 9 wounded on the cylinder, the secret message can get easily (Nag A., Singh J. P., Khan S., Ghosh S., Biswas S., Sarkar D. and Sarkar P., 2011). This method is called transposition method.

About 2000 years ago, cryptography was used by Roman army. Julius Ceasar was the commander of Roman army and he wanted to find a way of secure communication. A substitution method is also developed by Ceasar in which a letter is replaced by another symbol.

(16)

7 In this method, letter was shifted by fixed position or changed by other symbol and Ceasar took a lot of advantage of this method during the war (Senthil K., Prasanthi K. and Rajaram R., 2013).

In 1500’s, Bliase De Vigenere developed a new cryptosystem which was based on Alberti’s cipher disk. In Vigenere method, a square was used in which there are 26 alphabets of English language was on both axis, X and Y axis. If someone wanted to encrypt the message then one’s had to find the corresponding letter when finding plaintext on row and key alphabets on column (Senthil K., Prasanthi K. and Rajaram R., 2013).The fig. 1.2 is the Vigenere table which employing the encription.

(17)

8

1.5 Steganography and Cryptography

In ancient Greek writings the meaning of steganography is the name given to the science of hiding information visibility. Nowadays the biggest misconceptions are confused with cryptography in Steganography.

As well know cryptography and Steganography have different purposes. Cryptography only hides the content or meaning or both of to protect secret messages from unintended targets. Steganography quit different hides even the mere existence of the message. Thus Steganography gives more information security as also enhances privacy as confidentiality better than cryptography since it hides the existence of the message. Consequently, one basic disadvantage of cryptography systems is that the message still exists although encrypted.

Although, cryptographic and steganographic systems enhances confidentiality during communications, they process varying definitions in term of system breakings. For a cryptography systems, it is regarded as broken when an adversary can read the confidential message. While a Steganographic system is regarded as broken when an adversary can discover or identity the presence or existence or read the content of the message. Furthermore, a steganographic system can be regarded as broken if an adversary merely suspects a steganographic method or even a file without decrypting or deciphering the information. Because of this, cryptographic systems are considered less fragile than steganographic systems when it came to system failure. More ever Steganography systems must evade any form suspicion in order not to be considered as unsuccessful systems. As Steganography provides and additional cover of security to cryptography; combining steganography and cryptography results maximum confidentiality in private communication. Consequently, steganography aims to perfect cryptography and to evade the suspicions of adversaries, but not to replace cryptography completely (Almohammad A., 2010).

1.6 Steganography and Watermark

The purpose of the Steganography is to conceal the mere existence of the communication by covering or concealing the messages with other objects. On other hand, the purpose of the watermarking is to secure and protect the legal entitlements of products. For instance digital media like video, music, images and so on. It saved as a proof of originality outright coping or

(18)

9 small scale modification to watermarked file can be detected by the owner. Therefore, both of Steganography and watermarking are varies method of concealing messages.

Watermarking is a method that is used to hide data in order to prevent the removal of copyright information from digital documents. Even if is while watermarking exists in an object, attempting to remove the watermark would definitely cause a distortion of the original watermark. However, the kind of this watermarked document is called “robustness” characteristics of watermarking. On the basis of concealed or embedded data, there are two basic approaches of document marking: these are watermarking and fingerprinting. Whereas watermarking is an approach involves concealing a particular copyright mark into digital documents. Also the fingerprinting is a process by which a serial number is embedded in every copy of the digital document so that any break of licensing agreement can be detected.

1.7 Prisoner’s Problem

To further understand the concept of steganography quite clearly, it is imperative to understand the problem of prisoners proposed by the researcher (Simmons G., 1983) in which the illustration include two prison inmaters Alice and Bob who are serving prison sentences in different rooms or places with Wendy as an observer on them and as an intermediary responsible for the exchange as confidential as private communications between them, without knowing the content or making any effort to know the content or alterrpting to change the content before transmitting.

(19)

10

2. RESEARCH AND BACKGROUND

2.1 General Principles of Steganography

Steganography is arts and science of concealing secret message in communications over a public channel such a way that does not attract attention the existence of the message to any viewers who are neither the sender nor the intended recipient (Samphaiboon N., 2009).

To ensure the security of data as they travel within the network what kind of techniques must be utilized? Plus, in what manner would we be able to shield these information from recognition, vandalism or removable? Notwithstanding that, how this information get to their target or recipient without diminishment or deformity? Also what is the motivation behind the seeking of alternative approaches or seeking to complement old methods of securing data?

The procedure of encryption or encoding systems is used to ensure the security of the data which in light of the utilization of symbols, in addition to codes, and the utilization of complex mathematical approaches for changing the information into another picture that is hard to pursue or read, except by the concerned beneficiaries. Furthermore, encoding techniques has been utilized for a long time and till now in spy and military operations and in recent years being utilized to secure competitive advantages in commerce and exchanging information safely over the Internet. Notwithstanding what is supply by the encode techniques in the field of data security, however it has not offered the obliged security standard of the information, there are also individuals who are searching for routes harder to crack and hack encryption. Additionally, in today’s world, there are people who search for techniques to crack encodes, since the text can recognize it from the encoded or decoded text, making it makes simple for hackers to detect the behavior through direct incessant tests to get the encoded information, in order that, the utilization of encryption in a particular manner over and over to assist assailants (adversary) expel by reiterated tests, which was appealed to start search for another techniques more pliable security and conceal in the information, while the approaches utilized in this field, which is the topic of our study in this dissertation is steganography is an old system, yet it corresponded with the quick advance of interchanges used to conceal data inside other information.

(20)

11

2.1.1 Utilizes and goal of steganography

Steganography is the traditional method used in information hiding. The use of systems like cryptography it makes the output of these methods very conspicuous. The main reason behind this is that the systems make messages unreadable to anyone except the intended recipient and does not hide its existence. The design of steganographic approaches is aimed at hiding the message or (data) and ensuring that its existence is hidden from observers. To further buttress the security of the hidden message, the sender may encode it before it is hidden. This is beneficial since in some countries possessing encoded data is against low. The steganography enables this material to be concealed. In the contemporary world, steganography is used in watermarking. Watermarking are labelers that are put into material, this can then be used authenticate the data or for identification purposes. This may include, for example, submitting reviewers an audio file of unreleased song. In this case, the unreleased song bears the identity of the reviewer and if anybody leaks it in the internet the record label will simply have to get one of illegal the leaked songs, retrieve the watermark and whoever leaked the song will be identified. However, steganography is not all watertight technology for instance terrorists and criminals are increasingly using this method to hide data and their communications. Criminals perpetrating child pornography can use Steganography to conceal pictures in the photos on a legal eboy list and supply them to their clients. Al Qaeda has also been known to use this method to store data found in text files concealed in videos (Gallagher S., 2012).

Despite the threat posed to steganography by criminals and terrorists, this method can also be of great benefit to home users who can use it to store in a secure way, personal and private data. For instance bank accounts can be embedded in an image format. However this may provide to be insecure if somebody accesses the computer (such people may recognize an encrypted file and break it).

According to findings, some governments are using steganography to promote illegal and legal exchange of information. For example some Russian spies were arrested and challenged in US courts for using steganographic information channels particularly image Steganography to spy on the US and pass information to Russia (United States Department of Justice, 2015).

(21)

12 Suggestions from different quarters have proposed that steganography be adapted for transmitting information within a botnet. Nagaraja et al.(Nagaraja S., Piyawongwisal P., Houmansadr A., Singh V. Agarwal P. and Borisov N., 2014) proposed stegbot. Stegobot is botnet that employs Steganography as the basis for its command and control(C&C) network. Stegobot used one of the image formats like JPEG steganography to embed personal data such as credit card information and title deeds, passwords in images that are shared on facebook. Bots are linked to one another through users on the ancient. After the images are shared on facebook, they can be seen by all users who are connected to the uploading user.

2.1.2 Types of steganography

Steganography is one of the information hiding techniques which can be categorized into four parts, text, image, video and audio as show in Fig.2.1

Figure.2.1: categories of Steganography Steganography

Image Video Audio Text

Structural Random & Statistical Generation Linguistic

(22)

13

2.1.2.1 Text steganography

Text steganography can be classified into three main groups: Structural, random and statistical generational, and finally linguistic. The main characteristics of structural text steganography is that it modifies the physical form of the text, for instance through the appending white spaces and linespaces. On the other hand, random and statistical generation entails providing the cover text. This can be in a random manner or depending on specified input. While linguistic steganography uses contents of natural language, for example verbs, nouns, adjectives and so on. Linguistic steganography are classified into two: syntactic and semantic syntactic text steganography deals with changing the format of the text without considerable alteration to the meaning or tone (Gardiner J., 2012).

2.1.2.2 Image steganography

Currently image steganography is the most used cover media in steganography. It can be grouped either in image domain (spatial domain technique) or transform domain (frequency domain technique). Image domain basically entails concealing information in the list significant bits at specific intervals in the image. On the other hand, transform domain technique is used on the cover media to hide information in the lossy part of a compression algorithm. For instance, JPEG compression entails capitalizing on the discrete cosine transform. This makes information less susceptible to further compression algorithm (Morkel T., Eloff J., and Olivier M., 2005). Finally, fig. 2.2 contains secret message which hidden inside the image.

(23)

14 Figure.2.2: Example of image steganography

2.1.2.3 Audio steganography

Audio also plays on an important role in hiding data. This method majorly involves malling the hidden data. Not audible to the human ear. For instance, under normal circumstances, where there is a very loud tone followed a less loud tone. The human ears will not hear this second tone. So, this method can be used in hiding the data to replace the most significant bits too compress the audio. Another way it to apply a minute echo to sounds to hide data with the delays extending the waves on the sounds (England J., 2015).

2.1.2.4 Video steganography

Video steganography can also be utilized to hide data in every single frame. For instance, in one video obtained on from laptop belonging to an Al Qaeda suspect, it initially looked like a pornographic video but after forensic investigations it was discovered that the video had 141 hidden text files on operations and planning of the group’s terrorist activities (Gallagher S., 2012).

2.1.3 Types of text steganography

Generally as highlighted before steganography can be divided into four main parts which are: Image, text, Audio and finally video steganography. According to the cover media is utilized to conceal secret data. The sub rood of Steganography which is called text steganography can be

(24)

15 employs everything from modifying the format of current text, to changing words inside a text, to produce random character sequences or utilizing context-free grammars to produce readable texts (Bennett K., 2004).

To understand well, what mention fig.2.3 are types of text steganography; text steganography can be classified in three basic categories: structural based random and statistical generation and linguistic method. These three main forms of text steganography will be described here.

Figure 2.3: Basic categories of text steganography

In this section it is focused in detail of text steganography. Text steganography are divided into three groups which are: Structural base, linguistic and random and statistical generation.

2.1.3.1 Structural based

Text steganography based on structure involves controlling the structure or the physical form of the text for hiding data. Additionally, a characteristic of structural text steganography is that it modifies the physical form of the text, for instance through the appending white spaces and linespaces. Generally, structure based methods employs modifying current text in order to conceal secret message by utilizing non-displayed character, resizing the font and misspellings (Bennett K., 2004).

2.1.3.2 Random and statistical generation

Cover medium to hide data can be created by generating a variety of methods. It can be created statistically in which the chosen words match same statistical criteria. It can also be done randomly in which the chosen words are exclusively based on the information to be concealed.

Text Steganography

(25)

16

2.1.3.3 Linguistic

Linguistic steganography is a method which is based on the changing the meaning of the text is limited. Whatever, this method is divided in two forms: syntactic and semantic. Syntactic text steganography includes modifying the structure of the content without altogether adjusting the importance. For instance, appending punctuation in sentences. Semantic based steganography is the method that uses the synonym of certain words thereby hiding information in the text. However this method involves substituting words with their synonym (Gardiner J., 2012).

2.2 Steganalysis

Steganalysis corresponds to steganography or cryptanalysis corresponds to cryptography. The basic difference between is that cryptography deals with studying of analyzing information systems in order to deduce or determine the original data from an encrypted message. Steganalysis sequentially used to detect the presence of the hidden data in a message although in some instances it can be used to deduce the message.

2.2.1 Attacker models

The adversary is the outsider or the third party of the communication process by utilizing the Steganography. Eve who will be the observer on Alice and Bob to avoid them away from runaway, and reporting each other about the secret plans and forbidding the message from accepting it or analyzing it for knowing its substance or sending it after saving. For that reason when we are in Steganography, attackers can either be passive or active (Seidan Y., 2013).

2.2.2 Passive Adversary

When Eve represents the passive observer, he will capture messages sent among Alice and Bob. On other hand, they will be able to pursue all messages and Eve will not make any modification to the messages and will send them on in place to the collector. They will run investigation on the message to attempt and find concealed information. In the situation of the prisoner’s problem, this will be the superintendent just perusing the messages (Gardiner J., 2012).

(26)

17

2.2.3 Active adversary

When Eve represents the active observer, he will again capture messages sent among Alice and Bob, yet, this point will modify the message in order to remove any possible hidden data. For instance, if the hidden data was text steganography and specially structured steganography method, they utilize by clearing or removing the line spaces and white spaces to harm the hidden information, and also, if the hidden was an image or audio they could utilize to hurt the concealed data by execute the compression algorithms or noise reduction algorithms.

2.3 Text Steganography Techniques

2.3.1 Text abbreviation or acronym

Another important method employed to conceal secret information in a text involves the use of abbreviation or acronyms. In this method, very small amount of information can be concealed in the text. For example, it is only possible to hide a few bits of secret information in file of several kilobytes. A list of acronyms and their respective translation are presented in table

1 (Roy S., 2011).

Table 1: Shows the list of acronyms (Rafat K., 2009)

No. Acronym Translation

1. F2F Face to face

2. CM Call Me

3. C See

4. 218 Too late

5. ASAP As Soon As Possible

In this method abbreviations can be used instead of the complete words in order to represent the binary bit pattern of zero or one corresponding to the bits of secret information.

2.3.2 Semantic Method

This method employs a system where the information in the text is hidden by using synonymy of specific words the synonymy substitution may typify a single or multiple bit combination for the information intended to be kept hidden. One of the main benefits of this method is the protection of the information in case of retyping or using OCR programs.

(27)

18 Nevertheless, this method may change the meaning of text. A list of words and their synonyms are presented in Table 2(Roy S., 2011).

Table 2: shows the synonymy (Roy S., 2011)

No. Word Synonymy

1. Big Large 2. Chilly Cool 3. Small Little 4. Smart Clever 5. Spaced Stretched 2.3.3 Syntactic Method

Information in text file can be hidden by using punctuation marks and placing them in the right positions. In order to use this method, the suitable places for putting the punctuation marks, such as full stop and comma, should be determined. What is more, the amount of information to hide is of little significance or value of this method (Roy S., 2011).

2.3.4 Change of Spelling

This method takes advantage of the spelling difference between British and American English for concealing secret information bits. Table 3 presents a list of words that have different spellings in British and American English.

Table 3: shows the difference between American and British English (Roy S., 2011)

No. American Spelling British Spelling

1. Center Centre 2. Fulfill Fulfil 3. Favorite Favourite 4. Criticize Criticise 5. Dialog Dialogue 6. Check Cheque 7. Tire Tyre 8. Defense Defence

The methods that change the format of the text usually have a large of capacity for hiding information some examples of these methods are follows:

(28)

19

2.3.5 Line Shifting

In lines shifting method the text lines are vertically shifted to some degree in order to encode the document in a unique manner. This method is mainly used in printed texts. However, this method the distances can be observed using special tools of distance assessment and required changes can be done to eliminate the secret information. Moreover, in cases where the text is retyped or optical character recognition programs such as OCR are used, the secret information would get destroyed (Roy S., 2011).

2.3.6 Word Shifting

In word shifting method, secret information is concealed in the text by horizontally shifting the word and changing the distance between them. This method can be applied when the space between adjacent words is varying. This method is hardly identified because changing the distances between the words in order to fill a line is very normal. However, if one is familiar with algorithm of distance, he/ she can compare the text with algorithm and identify the secret information hidden in the text by using the difference. Moreover, the image of the text can be closely studied to determine the changes in the distance between words. Although this method is tedious, there is a high chance of identifying information concealed in the text. It is worth noting that using character recognition programs or retyping the text destroys the secret information hidden in the text (Roy S., 2011).

2.3.7 Open Spaces or White Spaces Methods

In this method, secret information is hidden by incorporating additional white spaces in the text. As showing in fig. 2.4, the white spaces can be placed at many different locations such as at the end of each sentence or paragraph. This method can be applied in any random text without raising the reader’s suspicion. However, the amount of secret information that can be hidden by using this method is very limited. Furthermore, same text editor programs tend to automatically delete and destroy the secret information.

(29)

20

Figure 2.4: Before encoding (Roy S., 2011).

Figure 2.5: Encode text (Roy S., 2011).

2.3.8 Steganography of information in specific characters in words

Information can be hidden by choosing characters in particular words. For example, the first words of each sentence are chosen in such a way that the hidden information can be determined by placing the first characters of the selected words next to one another. This method is believed to have been used by classic poets of Iran. This method is very tedious and requires strong mental power and special text (Roy S., 2011).

(30)

21

2.3 Current Research

There have been many attempts to develop methods of text steganography in different languages such as English, Arabic and Chinese. This section describes some of the work done on the text steganography. Following is the list of different methods of the work carried out and reported so far.

From the outset, below is an example of one of the most significant null cipher messages. Let us examine the following message sent by a German spy during World War II:

“Apparently neutral’s protest is thoroughly discounted and ignored. Isman hard hit. Blockade issue affects pretext for embargo on by-products, ejecting susets and vegetable oils.”.

As highlighted earlier, invisible message produced as a result of stego-text steganography, inspectors should not be able to arouse suspicion against contained confidential message.

The above stego-text can be decoded by extracting the second letter in each word to produce the hidden message:

“Pershing sails from NY June”(Yeh and Hwang, 2001)

Roslan N., Mahmod R. andUdzir N. in their documentation that deals with the text steganography in Arabic text, Arabic text steganography uses a sharp-edges method, which allows one to hide the secret bits in the sharp-edges for each character in the Arabic text document. This technology is very much dedicated to the hiding capacity of information. This technique allows hiding the information more in amount and hence very much needed that it is introduced to public also (Bhattacharyya D., Haveliya A. and Kim T., 2014)( Roslan N., Mahmod R. and Udzir N., 2011).

Shirali-Shahreza M., in his paper proposed a new method which involves, changing words spelling of American and British language to hide the information. For example there are many words in both languages which are spelled differently but pronounced same, simply the word “Favorite” is an American word but “Favourite” is a British word, however having these type of words with similar sound and different in spelling is an efficient way of information hiding (Shirali-Shahreza M., 2008a).

(31)

22 Shirali-Shahreza M., proposed a new method of synonym-based text steganography which makes use of the differences between American English and British English, for example the difference between the American word “Movie” and British word “film”. Using this mechanism the Author has proposed this method for hiding data in an English Text by substituting such words and hide bits.

In addition, this method converts the interested data to an arrangement of bit 0 and bit 1. The method will place a US word in a sentence for hiding of the bit 0 and will place the UK word in the sentence in order to hide the bit 1. This way, the data will be hidden in the interested text (Shirali-Shahreza M.,2008b).

Liu M.,Guo Y. And Zhou L. Proposed a text steganography method to be used in online chat. This method is based on an internet meme named typoglecymia, which allow changing the order of word’s middle letters and has a tiny to no effect on the ability of skilled readers to understand the text. For this reason, it used the redundancy found in the interior letter’s order (Liu M.,Guo Y. And Zhou L. ,2009).

Another method for hiding information is the use of abbreviations. So M. Hassan Shirali-shahreza proposed new method for secret exchange of information through chat by using abbreviation text steganography with the use of the SMS-Textting language. In this method, very little information can be hidden in the text. This table shows the list of acronyms in the left column and its transformation in the right column (Shirali-Shahreza M., 2007).

Gutub A., and Fattani M., proposed a novel Arabic Text steganography Method Using Letter Points and Extensions. The approach hides secret information bits within the letters benefiting from their inherited points. To note the specific letters holding secret bits, the scheme considers the two features, the existence of the points in the letters and the redundant Arabic extension character. The authors use the pointed letters with extension to hold the secret bit ‘one’ and the un-pointed letters with extension to hold ‘zero’. Not all letters are holding secret bits since the secret information needs to fit in accordance with the cover-text letters. Redundant Arabic extension characters are used beside the letters to note the specific letters holding the hidden secret bits. The nice thing about letter extension is that it doesn’t have any effect to the written content (Gutub A., and Fattani M., 2007).

(32)

23 1995 Maher K., has proposed a data hiding program called TEXTO. TEXTO is a text steganography program which is a particularly well suited way of exchanging binary, encrypted data. TEXTO is designed to transform unencoded or PGP ASCII-armored ASCII data into English sentences. Maher’s method works just like a simple substitution cipher, in which the secret data is replaced by English words. But verbs, nouns, adjectives, and adverbs are used to fill in the preset sentence forms or frames, also these words are important in the resulting text. However words such as connecting words and punctuations or any words not recorded in the dictionary are ignored.

The first incompetence of this method is that the created cover text perceptually does not to appear. The second one is the lack of semantic integrity of the text. Therefore, the communication between the parties is the reason that employees observes to analyze and create suspicious situation.

Chapman and Davida, have proposed two methods called NICE TEXT and SCAMBLED which consists of two functions. NICETEXT, uses a portion of the text to embed a message in the form of changing synonym place (Desoky, 2009). For the last 15 years, Synonym based approaches have attracted the attention of many researchers such as Winstein (Wintesin, 1999), Nakagawa (Nakagawa and ark., 2001) and Murphy (Murphy and Vogel, 2007). Synonym based text steganography must protect and change the meaning of cover text. When choosing synonym words to function in an appropriate manner, with regard to stego text, the words should be reasonable in terms of the rules of language.

In 2009, Desoky proposed linguistic text steganography method called Listega which uses textual list of items to camouflage data by employing itemized data to hide messages. Clearly, Listega first converts a message into binary then assigns it to appropriate items in order to generate a cover text in a structure of list. Listega obtains genuine message by basing on camouflaging both the message and its transmittal on an appropriate or legally list of items.

Evidently, such steganographic cover text in a structure of list of items is logically and linguistically legitimate. Additionally, Listega uses combinatorics based coding in order to provide the preferred randomness while the parts are communicating. Combinatorics coding is predictable to the receiver but quite random to an observer (Desoky A., 2009).

(33)

24 Lee and Tsai (2010) proposed a new approach for concealing secret messages in PDF file as cover communication. The proposed method specifically uses two techniques of data encoding, alternative or substitute space coding and null space coding, in the encoding approach they used special ASCII code A0 for message embedding in the texts of PDF documents.

In the proposed method, a secret message is considered as a string of characters or bits and encoded with a special ASCII code by unitary or binary coding. As highlighted before, in the two approaches or techniques, the secret message were concealed respectively between words or between characters and became invisible in the general PDF reader windows.

Simply, for using between-word embedding and between-character embedding, the width of A0 in between word embedding was set to be similar to that of the space code 20, and the width in between character embedding was set to be zero. In the concealing results in this method and resulting stego-PDF file were established to be invisible in the windows of common PDF reader and establishing very successful Steganography and achieving the purpose of secret communication (Lee and Tsai,2010).

The proposed method’s capacity depends on a number of characters in the PDF document which is desired as cover media. Besides that, low security which means, if an observer knows the algorithm can detect the hidden information by applying the inverse of concealing procedure.

Shu and Ark, (2011) proposed an alternative-text information hiding algorithm, whose cover text controls the process of hiding information. The secret message would be concealed successfully according to the amount and the type of the cover to be diffused into multiple text segments to be stored. In the Author’s method, the secret message would be divided into 2n parts by XOR analyzing of n (n>=02) times by according to the number of cover texts and the cover text must be equal or more 2n. After that, every part of the secret message is concealed into different cover texts by depending on categories of cover texts.

For example, when the cover text concerns technology and mathematical science with mathematical codes, punctuation marks and too much mathematical formulas, the method of embedding between the words categories could be embraced. Also, when the cover text is about literature, the method of transferring the cover information, such as synonym substitution, could be adopted to embed the secret message. At the same time, when the cover text is normal text, an illustrated text or cleared text, the method of employing the format of cover text could be

(34)

25 adopted. However, the key could be utilized to conceal the hidden information by the other types of cover texts. Here, the corresponding relationship between text segments is part of the key of the hiding algorithm.

On the other hand, only receiver and sender know the relation; therefore, it is extremely complex for an adversary or any unauthorized persons to get it (Shu and ark., 2011). Adopting more carrier texts and defining the concealing procedure according to the category of cover text, in order to embed secret message, makes the stego cover more soft against the adversary. However, the disadvantage of algorithm can be uncertain or questionable while the algorithm is known to the public, if no key is adopted.

In 2011, Ryabko and Ryabko proposed a system in which the probability of distribution of carrier texts with and without secret information are equal, because the carrier texts are generated by a finite-memory with possible unknown statistics. For that reason, the Ryabko’s Steganography systems are completely very secure. However, an observer cannot decide whether secret or hidden message is being transferred or not.

The main and essential idea for the system that was proposed by Ryabko is the following: presume that for a carrier text X created by a path or source, then it can find a set S of carrier texts, after that every carrier texts in S have equal probability of being generated as X. Besides, suppose that every element of S determines S uniquely. Then, in the place of sending the carrier text X that was really created or generated, it can be transfer the carrier text in the set S which number in S is compatible to the hiding message that is wanted to be passing. This does not change the probabilistic qualities of the source, gave the secret message comprises of (i.i.d) equally probable bits. In this way; an adversary can't tell whether secret message is being passed. There are two disadvantages of the sketched out stegosystem: first and foremost, the rate of transmission of secret message is not ideal, and second, it applies just to i.i.d. carrier texts (Ryabko and Ryabko, 2011).

Since the interactions by the use of social services or chat room has become more prevalent in individual’s live, in 2009, Wang and Chang proposed another text steganography method that conceals secret information into emotional or image icons in chat rooms through the internet. In this technique, firstly, the sender’s emoticons table ought to be consistent with receiver’s emoticons table. This means every item in sender’s table must be similar to the ones in

(35)

26 the receiver’s table. After that, the sender and receiver categorize those emoticons according to shared emoticon table into several groups by their meaning and each emoticon belongs to group that have the similar meaning, for example “smile” emoticon group, and “cry” emoticon group, and “laugh” emoticon group, and so on.

As highlighted before, emoticons are categorized, and every emoticon has an order number which counting from 0, in its group is the secret bits that will be concealed. Therefore, the proposed method utilizes a secret key to manage the order of emoticons in every constructed set.

The authorized parties who can keep this key are the sender and the receiver. While the concealing capacity was enhanced because of great numbers of emoticons utilized in many kinds of chat rooms, it generally relies upon already pre-shared or preprocessed emoticon table and number of items in every emoticon set (Wang and Chang, 2009b).

In (1992, 2002) Wayner, presented a new approach of steganography which is called mimic functions. In Wayne’s new approach, input data steam of randomly distributed bits and the inverse of Huffman coding is applied. The purpose of this process is to produce a stego text matching the statistical profile of a normal text. Therefore, the text produced by the stego mimic functions, is resistant to statistical attacks. To enhance the output of mimic functions, it can utilize the concept of context grammars and vanwijaarden grammars. Actually, the result of a normal mimic functions is hogwash. Appropriately, the gibberish makes the content amazingly suspicious. On the other hand, utilizing mimic function with context-free grammar slightly enhanced the comprehensibility of the text. But yet, the cover-text still contains various defects such as syntax mistake, grammar and lexicon error, and incorrect rhetoric. So that, the contents of the cover-text, which allows us to hide the information will be meaningless and semantically disconnected. These weaknesses may bring suspicion up in covert communications (Desokey A., 2009).

Sun et al (2004), proposed an approach that uses the left and right elements of Chinese alphabet characters. The name of the proposed scheme was L-R approach. The proposed method utilizes the mathematical statements of all Chinese characters into the text concealing procedure. It picks those characters with left and right components as candidates to hide the secret message.

(36)

27 During the embedding stage, there are two steps adopted to embed the secret information in this scheme. As highlighted in the previous lines, Chinese characters can be constructed left-and-right components that would be picked to hide the message. If the secret message that wanted to be embedded is 0 bit, the proposed scheme must be divided into left and right component of the Chinese charters. But if embedded bit is 1, the proposed scheme does not need to divide the candidate character into left and right components. The second step is to adjust the spaces between left and right components of the current candidate character to keep the candidate characters’ original quality. In the extraction stage, the proposed scheme computes the spaces and widths between neighboring characters to illustrate whether it is hiding the secret message or not.

In 2009 Wang et al. reviewed the sun’s scheme and proposed enhanced scheme which incorporates L-R scheme to the up and down structure components of Chinese characters instead of only the left and right components. Moreover, the enhanced scheme acquired reversibility, which allows the original cover-text to be easily extracted and recovered after concealing the secret from the stego-text, so that the recovered file can be utilized over and over again for the later cover transmissions. Also, the extracting strategy is simplified by omitting the need to use an image in the extracting procedure.

A new approach which is called Translation-based steganographic scheme has been proposed by Grothoff et al. This approach conceals secrets in the noises which are characteristically facing in a machine translation (MT). It could be embedded the secret information by carry outing substitution stage on the translated text, utilizing translation variation of multiple machine translation systems. Moreover, it adds popular noises of MT systems and as well, synonym substitutions utilized here in state to increase or enhance the bit rate. Unlike synonyms-based steganography, linguistic flaws in noise-based approach are not disturbing if they don’t appear excessively. Besides, Grothoff et al. illustrates that; the continual improvement of MT can be one of the issues that narrow the margin of hiding data. In addition, as highlighted Grothoff et al, the proposed method which is named translation-base cannot be utilized to all language to transform because of the basic structures are fundamentally different. So, this produces robustly incoherent and incomprehensible text.

(37)

28 Another noise-based approach was proposed by Topkara et al. in 2007. Here, typos and ungrammatical abbreviations in a text, e.g., emails, blogs, forums, etc., are employed for hiding data. The lack of these approaches is that they are sensitive to the amount of noise that occurs in a human writing. This both increases the vulnerability of the approach and also narrows the margin of hiding data (Desoky A, 2009).

Samphaiboon (2009), developed Steganography scheme that covertly sends secret information to different collectors through a running stream of short text messages showed on a media (for example websites television sets). In this approach, it expected that the optical character recognition (OCR) usefulness was accessible at the decoder stage. Also, as case study, was utilized Thai language short text messages and analyzed the features of Thai short text messages and acquainted some successful message to bit transformation method in concealing process. Moreover, the author was established in his analyze in, one Thai short text message can be transformed into multiple secret information bits which means, it can be send more bit secret message in the every cover. Besides that, the proposed method can be utilized in any language to transform short text communications. In an empirical evaluation, illustrated that 4 secret information bits can be concealed in every short text message. Also, they cleared that the extraction process can be simply and correctly classified the concealed secret from the cover by manually rather than utilizing OCR functionally. So, the method is useful for transforming secret information from one sender to more than one collector in different places at the equal time through public channel. Other hand, the author believes the control of the short text message channel is under the sender (Samphiaboon N., 2009).

In 2012 Satir and Isik proposed compression based text steganography approach which determines the most significant parts of Steganography which are capacity of secret message and security of hidden data. First, the compression algorithm which is named LZW has good performance and employs to expand the hidden message’s limit. Also the second technique which is used in this algorithm is combinatoric-based coding(Latin square) which increases the security and supply as randomness for selecting camouflage techniques or email id’s to conceal the secret message with the prepared text base and utilized it forward mail platform as stego-media in order to conceal the secret message. The experimental result gets that the capacity can

(38)

29 be up to 7.042% for secret information which contains 300 characters. On other hand, if its capacity compared with previous methods the author’s capacity is superior. Beside, that concealed data or message cannot be extracted simply if not used steg-key (Satir E. and Isik H., 2012).

Also (Satir and Isak) proposed another method, which uses the same procedure as previous method, but different from the way of compression. This type of compression is called Huffman, Huffman compression based text steganography was proposed after the author seen the LZW’s capacity was not comfortable to the algorithm’s goal which was creating imperceptibility and increase the security when the receiver makes the extraction of secret message. On other hand, the objectives of this study is to focus on an important increment ration in terms of capacity and more complicating the extraction way for the security of the hidden information. Besides, the purpose of complicating the extraction procedure the concealed data has been camouflaged in the text chosen from the already developed text base. The cover media or text base of the posed methods included in, article’s abstracts and notification texts and so on, so, this natural texts are meaningful and the syntax, grammar are correct, and it also reasonable. Also there is another technique used in this algorithm and named combinatoric-based coding (Latin square) which increases the security and supply as randomness for selecting camouflage techniques or email id’s to conceal the secret message with the prepared text base and utilized it forward mail platform as stego-media in order to conceal the secret message (Satir E. and Isak H., 2012).

.

In 2014 Ahmad T., Marbun M., Studiawan H., Wibisono W., and Ijtihadie R., proposed novel random email based text steganography , this method is extracted from (Satir E. and Isak H. 2012) and it makes some improvement such as the random email addresses of the second stego key of the Satir and Isak. However this random can increase the security of the hidden data and the length and generating of the cover text establish the successfulness of the concealing stage. By the way, the probable cause of the fail of embedding process is the non-changed and short cover text (Ahmad T., Marbun M., Studiawan H., Wibisono W., and Ijtihadie R., 2014).

New approach for text based steganography

TABLE OF CONTENTS