Magic Random Bits LSB Audio Steganography

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/288824542Magic Random Bits LSB Audio SteganographyConference Paper · March 2015CITATION READS1 1766 authors, including: Abdul Hakeem Hazara University Zakir Khan 7 PUBLICATIONS 2 CITATIONS Hazara University 28 PUBLICATIONS 24 CITATIONS SEE PROFILE SEE PROFILE Noor Ul Amin Hazara University Mohsin Shah 32 PUBLICATIONS 74 CITATIONS Hazara University 17 PUBLICATIONS 26 CITATIONS SEE PROFILE SEE PROFILESome of the authors of this publication are also working on these related projects: Ensuring efficient security mechanisms using parallel computing View project Stegangraphy View projectAll content following this page was uploaded by Zakir Khan on 06 December 2016.The user has requested enhancement of the downloaded file.

Magic Random Bits LSB Audio Steganography Abdul Hakeem MSCS Mohsin ShahDepartment of Information Technology Department of Information Technology Hazara University Hazara University Mansehra, Pakistan Mansehra, Pakistan [email protected] [email protected] Zakir Khan MSCS Abdul Qadir MSCSDepartment of Information Technology Department of Information Technology Hazara University Hazara University Mansehra, Pakistan Mansehra, Pakistan [email protected] [email protected] Adnan Ahmad BS.C Noor Ul Amin Department of Electrical Engineering Department of Information Technology University of Engineering and Technology Hazara University Mansehra, Pakistan Peshawar, Pakistan [email protected] [email protected] the word “Steganos” which means covered, concealed orAbstract—Audio steganography is the technique of hiding secret protected. Steganography is the branch of science in which ainformation into the samples of an audio signal. In this work a message is embedded on transmission side and is retrieved onnew least significant bit (LSB) audio steganographic technique receiver side. It may be to prevent piracy for copyrightis introduced. Proposed technique works on the basis of random protection or private personal communication. “Host” orbits steganography. These number of bits for steganography of “cover message” is used to hide user’s message. Thenth sample is be defined by (n-1)th sample’s 3LSB decimal combination of host message and user’s message is calledvalue. The proposed technique does not affect the quality of the Stego-messages which is seemed same as the cover messageresultant audio signal and the difference between original and or host file. [2] An audio steganographic comprises of thestego audio is unnoticeable as well as it is more secure than the following components:conventional LSB technique due to random amount of data insamples of audio.  Message signal (Secret information).  Cover signal (Audio signal in which secretKeywords: Audio Steganography, Least Significant Bit (LSB),Security, Random Bit, Secret Data information is to be embedded).  Encoding algorithm I. INTRODUCTION  Stego signal (The resultant audio signal after The revolutionary development in internet, digital information embedding).communication has created huge changes in this overall  Decoding algorithmculture. It is very easy and simple to use software and also thecheapest prices of CD and DVD recorders, Laptops, PDA’s The goal of steganography is communication take placehave played an important role to exchange data and for the secretly between two parties, without aware of the thirduse of multimedia. The error free transmission to distribute party.[3] Regarding audio Steganography there are four mainfiles between users are due to the broad band connections. In categories least significant bit coding, echo coding, phasethis advanced, enhanced and developed communication data coding and spread spectrum coding these are commonly usedhiding is an important factor for network security issue. to hide user information into auditory data. These fourDifferent techniques are used for sending important data via methods have their own implementation, bandwidth andinternet to keep them secret. In literature some methods are overtness and also each of them have advantages andused for data security e.g. watermarking, encryption and disadvantages.steganography. In this developing technology electroniccommunication has played a vital role because of its II. LEAST SIGNFICANT BIT ENCODINGsimplicity and security. While transmitting data it is veryimportant to make it secure. Steganography originates from In this technique least significant bit of each audio frame is used for embedding the secret binary information. This technique is very simple and having high embedded

payload in audio cover file. This is less secure technique A. Parity Codingbecause during file conversion secure data can destroy orcontaminate. This embedding technique relates with sample regions which are used for data embedding [8]. In thisA. Echo Hiding technique original signal splits into separate region of samples instead of individual samples. If the secret bit which is to be Echo hiding relates with the human auditory system. encoded doesn’t match with the sample regions of parity bitThis method is evaluated by creating and introducing short then process flick the LSB of one of the samples in the regionecho to the cover signal. During hiding of data following [9]. The advantage of this technique is that the dispatcher hasthree parameters related to echoes changes: 1.amplitude 2. more choices during encoding in the secret bit.Decay rate 3.and offset (delay time) from the original signal[5].Below human auditory threshold value all parameters in B. XORing Methodecho hiding do not work. For this we embed the data into thelast bit of original echo. The disadvantage of this technique In this embedding technique XOR operation isis that the embedding rate and security is low due to which performed. By XORing the data is selected on the basis ofup till now no new audio steganography system has been result after XORing and also on the bit which is to beintroduced on echo hiding. embedded, the LSB of the sample is customized or kept unchanged. The advantage of this technique is that thisB. Phase Coding increases the capability of the cover audio by 8 times greater and provides strong encryption [12]. In this embedding technique we modify the selectedcomponents called phase components which are taken from C. Bit Selectionthe original (cover) speech spectrum with embedded (hiddendata).This is true that human ear is most discernible to noise In this embedding technique we select different bitsas compared to the phase components of sound. This result to in each sample and then embed data in it .In this method theinaudible encoding in term of signal to professed noise ratio 1st three LSB’s and starting two MSB bits of a sample are(SPNR) in order to not detect by steganalysis method based used for data embedding and selection of bit of that sampleon SPNR, then the secret message gets masked in the audio respectively. The advantage of this technique is thesignal. The drawback of this technique is that due to the randomness of bits that’s why intruder cannot know the actualinsensitivity of Human Auditory System (HAS) to phase bit. [13].The 3rd LSB is used for hiding of data if starting twodistortion an attacker can easily ruin introducing MSB’s of the sample are 00 and so on.imperceptible frequency modulation. D. Sample SelectionC. Spread Spectrum In this embedding method only few samples are used In this embedding technique frequency spectrum of for hiding of data instead of using all the samples. The samplethe audio file is used on which secret data is scattered by number is created randomly which is proscribed by 1st threeusing a particular code which is not dependent on the actual (3) MSB’s. Let’s suppose current sample is i, then next sampleaudio signal [5].The last signal requires more bandwidth than will be in last column including secret message bit. Thereally required for transmission. There are 2 versions of number of samples skipped between two subsequent secretSpread Spectrum which can be used in audio steganography: message bits is equal to one more than 1st three bits (decimalthe direct sequence and other is frequency hopping schemes value) [13].[10]. Direct sequence spread spectrum describes aninvariable called the chip rate is used to spread out the secret E. Lowest Bit Codingmessage and then modulate with the pseudorandom signaland inter leaved with the cover signal. Frequency hopping In this embedding method least significant bit (LSB)Spread Spectrum relates, to hops rapidly between of the original message is used. This method also explains thefrequencies, the audio file’s frequency spectrum is changed embedding of wave data by lowest bit method. Both the data’s[11]. The drawback of this technique is that the noise and i.e. wave and secret data’s are in binary system and the lowestdistortion can be easily introduced. bit of wave data is exchanged with the secret data one by one [10]. The drawback of this method is that it results less change III. Various Data Embedding Techniques in LSB: and gives embedding capacity up to 12.5 percent of the wave file.Following techniques used for data hiding in LSB techniqueare: F. Variable Low Bit Coding This is one of the advance technique related to lowest bit coding method having high embedding capacity. Let data range of audio is from 0 to 255 and the middle range

is 128 at which silent of sound [10]. As a result in the middle A. Data Encodingrange data cannot be embedded by using this calculation thestandard level and two thresholds 1and 2.secret data cannot be In order to embed the data in an audio stream theembedded, if amplitude range is less than threshold 1. If current algorithm consists of the following steps. To embedamplitude range is between threshold 1 and 2 then for data data the algorithm reads the secret data that is to be embed,embedding only one bit is used and 2 bits are used for data then the algorithm proceeds to read the audio file in which theembedding if amplitude range is ahead of threshold 2. The data is to be embed. Then every single sample of the audio fileadvantage of this technique is that it has high embedding is read by the technique and next, are needs to read 3LSB fromcapacity it gives better SNR high security as well as the source audio sample. After reading the selected 3LSB wefrequency. are needs to calculate the decimal value of the 3LSBs. As the decimal value of the read 3LSBs is between 0-7 so theG. Average Amplitude Method embedding of data of the next sample depends upon the calculated decimal value. At the end of samples the stego Average amplitude data of neighboring audio data audio is written on the memory and then transmit toward theexcluding own audio data is used as threshold. If amplitude receiver this file is known as stego audio file that carries therank is greater than threshold, than two binary digits are used secret data. All steps are shown in the Fig 1.for embedding else no embedding take place. The advantageof this method is that this method will increase the capacity B. Data Extractionand secrecy of embedding. The disadvantage of this techniqueis that it is more complex and it is computational and also this To extract the data from the file, first needs to readmethod is limited to two binary digits. the stego audio of the source file. Then every sample of the stego audio is read. Just like embedding 3LSB’s are read and IV. PROPOSED METHOD then decimal values of these 3LSB is calculated. Then according to the decimal value of current sample the secret bits are extracted from the next sample. At the end all secret bits are concatenated together and written down as secret data file as Fig 2 Specify.FIG 1. DATA EMBEDDING IN AUDIO SAMPLE FIG 2. DATA EXTRACTION

FIG 3. MATLAB STANDARD AUDIO The quality of produced Stego audio is high and FIG 4. STEGO AUDIO humans cannot detect the changes between standard MATLAB audio and Stego audio. The difference between the FIG 5. DIFFERENCE BETWEEN ORIGINAL AND STEGO AUDIO standard audio and Stego audio is shown in figure 5. V. EXPERIMENTAL RESULT Results are shown on the basis of following parameters which are shown below in table 1. For the implementation of the proposed technique,we used standard Matlab audio generally known as Handel Results show that the proposed technique hides ashown in figure 3 with following specification high capacity of information in the standard audio with veryAudio Length = 8 sec minor changes.Original Audio Size= 142 KB VI. CONCLUSION MATLAB is used to implement the proposedtechnique for producing the Stego audio represent in Figure Random bit encoding technique is simple and4. efficient. The proposed technique does not affect the quality of the stego audio signal and is more secure than the TABLE 1: RESULTS conventional LSB techniques. This random bit technique do not affect the audio quality of Stego audio file and the Parameters Values embedded secret information can be recovered with the help of reverse steganography algorithm. As the random no for theMean Square Error 2.7290e-007 embedding are generated with 3LSB of samples so it has very low difference which is undetectable.Average Difference 5.9705e-006 ReferencesMaximum Difference 0.0038 [1]. J. Johnston and K. Brandenburg, 1992, \"Wideband CodingNormalized Absolute Error 0.0956 Perceptual Consideration for Speech and Music”. Advances in Speech Signal Processing, S. Furoi and M. Sondhi, Eds. NewPayload 65535 Bits York: Marcel Dekker.Audio Length 8 sec [2]. Asad, M. ; Telecommun. Eng. Dept., Univ. of Eng. & Technol. Taxila, Rawalpindi, Pakistan ; Gilani, J. ; Khalid, A.” AnOriginal Audio Size 142 KB enhanced least significant bit modification technique for audio Steganography”Stego Audio Size 142 KB [3]. Pooja P. Balgurgi, PG Student Elec. & Telecommunication department SKN College of Engineering Vadgaon Bk. Pune, India “Intelligent Processing : An Approach of Audio Steganography”. [4]. Ishaque, M. Qudus Khan, F. Abdul Sattar, S. Investigation of Steganal- ysis Algorithms for Multiple Cover Media.October 2011, Ubiquitous Computing and Communication Journal.Vol 6, no 5. [5]. F.Djebbar, B.Ayady, H.K.AbedMeraimx, 2011,” A view on latest audio steganography techniques”, International Conference on Innovations in Information Technology. [6]. P.K.Singh, R.K.Aggrawal, 2010,” Enhancement of LSB based Steganography for Hiding Image in Audio”, International Journal on Computer Science and Engineering, Vol. 02, No. 05. [7]. K.Geetha And P.VanithaMuthu, 2010, ” Implementation of ETAS (Embedding Text in Audio Signal) Model to Ensure Secrecy”, International Journal on Computer Science and Engineering, Vol. 02, No. 04. [8]. H.B.Kekre, A.Athawale, S.Rao, U.Athawale, October 2010 ” Information Hiding in Audio Signals” ,International Journal of Computer Applications, (0975 – 8887)Volume 7– No.9. [9]. M.Wakiyama, Y.Hidaka, K.Nozaki, 2010, ” An audio steganography by a low- bit coding method with wave files”, Sixth International Conference on Intelligent Information Hiding and Multimedia Signal Processing.

[10]. P.Dutta1, D.Bhattacharyya, and T.Kim, June 2009 ” Data Hiding in Audio Signal: A Review”, International Journal of Database Theory and Application,Vol. 2, No. 2. [11]. S.K.Bandyopadhyay, D.Bhattacharyya, D.Ganguly, S.Mukherjee and P.Das,” A Tutorial Review on Steganography”, 1Computer Science and Engineering Department, Heritage Institute of Technology, Anandapur, Kolkata – 700107. [12]. H.B.Kekre, ArchanaAthawale, SwarnalataRao, UttaraAthawale, October 2010, Information Hiding in Audio Signals, International Journal of Computer Applications (0975 – 8887) Volume 7– No.9. [13]. M.Asad, J.Gilani, A.Khalid, 2011, ” An Enhanced Least Significant Bit Modification Technique for Audio Steganography”, IEEE978-1-61284- 941-6/111$26.00 .View publication stats