INTRODUCTION

The World Wide Web, cognized as the inspiration for technological evolution in the epoch of wireless communication has revolutionised the digital world by revivifying the data rates a manifold. Coalesced with the evolution of internet, the number of users across globe reached a huge number but simultaneously several methods of hacking also increased. This in turn calls for higher data rate along with sound security. OFDM is a spectrally efficacious multichannel modulation scheme whose orthogonality has made it attractive over all modulation schemes. The orthogonality was first introduced by Chang (1996). Chang and Gibby (1968) typify that multiplexing can be done in OFDM. The parallel transmission in OFDM was described by Saltzberg (1967).

Higher data rates with effective spectral utilization was presented by Zimmerman and Kirsch (1967). Weinstein and Ebert (1971) suggested data transmission through Discrete Fourier Transmission (DFT). Peled and Ruiz (1980) introduced cyclic prefix to improve data transmission through frequency domain. A stego system requires cover image which holds the secret data (Kumar et al., 2011). The hidden data can be an image, bit, ciphertext or a plain-text (Katzenbeisser and Petitcolas, 2000; Xiang et al., 2011; Yang et al., 2011). The secret data embedded within the cover image forms the stego image (Praveenkumar et al., 2012a, b).

Robustness, security and capacity are the key features of any digital image steganography (Cheddad et al., 2010; Amirtharajan and Rayappan, 2012a-d; Amirtharajan et al., 2012; Zhao and Luo, 2012; Zhu et al., 2011). The number of bits that can be embedded decided the capacity of the cover image (Amirtharajan et al., 2012). Robustness refers to the resistivity of the image after data embedding (Marvel et al., 1999; Padmaa et al., 2011; Praveenkumar et al., 2012c). Security deals with the unintended user figuring out the secret data in the cover image (Amirtharajan and Rayappan, 2012a-d).

To prevent unauthorised access and copyright of the data in digital form, confidentiality and data integrity are essential as most of the data access are done through the internet, while to eradicate it, some secret data has been embedded (Thanikaiselvan et al., 2011a, b; Amirtharajan and Rayappan, 2012a-d; Amirtharajan et al., 2012) in the digital data which cannot be extracted easily without the intended algorithm. Cryptic effect can be introduced to the input image by applying any of the cryptographic algorithm and then data embedding was carried out to ensure two levels of security (Thenmozhi et al., 2012).

Aforementioned literature shows that, there are excellent works carried out in information hiding in image as cover both in time and frequency domain. Most of the implementation has been carried in matlab, Visual C++ and Java or in some programming languages and not even a few in simulink. Hence this study takes an optimistic effort to implement image steganography through OFDM in simulink. The succeeding section explains the proposed simulink model followed by results, discussion and conclusion.

PROPOSED SIMULINK MODEL DESCRIPTION

Read the cover image using Image from file block. To workspace block is used to display the pixels of the image in the workspace. Reshape is used to convert 2D array into 1D. Then integer into binary conversion is done using Integer to Bit converter and the number of bits per integer is 8. Reshape1 will convert single dimension binary bits into 8 rows and 65536 column matrix. By giving the starting and ending value of the required row and column to the Sub matrix extracting the required values from the matrix can be done.

Transposing rows to columns and vice versa is done using transpose block. To read the data from workspace constant block is used. Reshape 2 will convert the given data in to K rows and 65536 columns, where K denotes the number of bits to be embedded. The data Output is converted into 1D array and is further converted into integer values by Bit to integer block. Then the data to be embedded is displayed in the workspace using to workspace data_embedding block.

Matrix concatenate block is used to concatenate the outputs of the blocks. The output of this is given to the Transpose 2 block. Then the matrix obtained is reshaped into 1D array using Reshape block and the binary bits are converted into integers using Bit to integer block. Then the integer matrix is converted into [256 256] matrix using reshape 5 block and this matrix is displayed in the workspace. This is the pixel matrix in which the data has to be embedded.

The integer matrix is converted into binary matrix and is displayed in the workspace using to workspace 9 that will perform integer to bit conversion. Then the embedded image is transmitted through the OFDM sub system. The input to the OFDM system is the binary matrix. The subsystem contains BPSK/QPSK/QAM modulator, IFFT block, AWGN channel, FFT block and BPSK/QPSK/QAM demodulator. Reshape 4 is used for converting the 1D array of the OFDM subsystem output into a matrix of 8 rows and 65536 columns and is displayed in the workspace.

Submatrix1 is used to extract the required data matrix from Reshape 4 block. Then it is converted into 1D array and the bits are converted into integers and the received data has been displayed in the workspace using To workspace 8 block (received_data). Then the BER graph will be plotted using BPSK and QAM modulation techniques. Finally, the MSE and PSNR of the image will be computed for K = 1 and 2 bit embedding and the results will be tabulated. Flowchart of the proposed methodology is given in Fig. 1, subsystem of OFDM system and the overall simulink model of OFDM transmission and reception are given in Fig. 2 and 3, respectively.

Fig. 1:	Flowchart of the proposed methodology

Fig. 2:	Subsystem in OFDM Simulink model

Fig. 3:	Simulink Model of OFDM with data embedding, transmission and reception

Algorithm for data embedding and extraction: