Selective Encryption of CAVLC for H.264/AVC N. Khlif, T. Damak, F. Kammoun, N. Masmoudi Circuit and System Team "C&S", Laboratory of Electronics and Technology of Information, National School of Engineers of Sfax Sfax, Tunisia naziha.khlif@gmail.com, damak.taheni@gmail.com, fahmi_kammoun@yahoo.fr, nouri.masmoudi@enis.rnu.tn Abstract— From years ago, encryption became a necessity to transmit data securely. Like message written, fixed images or movies should be encrypted also while transmitted over a public channel. Currently, the literature is pointed to selective encryption for those types of data against attacks. That’s to say, the encryption module is inserted into the chain of video compression to reduce the computational time, ensure the security and the reduction of the amount of video data. This paper presents one method of selective encryption based on chaos for H.264/AVC video compression. It consists of the insertion of the encryption part in the entropy coding module (CAVLC: Context Adaptive Variable Length Coding). The results will be deducted according to the values of PSNR (Peak Signal to Noise Ratio), SSIM (Structural SIMilarity) and the rate of encryption. Selective Encryption CAVLC (SE-CAVLC) is interesting since it consider both Intra and Inter frame. Keywords— Selective encryption, H.264/AVC, CAVLC I. INTRODUCTION Transmitting data through a public channel is not benefic without ensuring their security. Certainly, for the transmission of fixed images and video, compression is required to reduce the volume of data in order to avoid clutter of public communications networks. So cryptography will be associated with compression to secure those data. The known standard encryption algorithms are unable to encrypt large amounts of data in real time especially for applications where time is a critical factor as the issue of satellite programs and telemedicine. For this reason, researchers have chosen the selective encryption. This approach means encrypting only a part of the data to reduce the computation time, ensure their security while transmission and maintaining a certain level of compression. In this paper we propose an encryption algorithm based on chaos inserted in the chain of H.264/AVC video compression on entropy coding module (CAVLC unit). This encryption concerns a certain quantified coefficients put in the entry of the CAVLC unit. This is for the three components of luma, blue and red chroma for both I and P frame. In addition, we present metrics to analyze the effect of encryption, such as PSNR, SSIM and the rate of encryption which represents the ratio between the number of bits encrypted and the entire bitstream. We present in Section II, the state of the art different methods of selective encryption of H.264/AVC and specially we describe the SE-CAVLC method. In Section III, we discuss the experimental results taken according to different analysis of the effect of encryption parameters. Finally, in Section IV, we conclude and we consider the perspectives of our work. II. STATE OF THE ART The video compression chain H.264/AVC [1] is a partnership effort between the ITU-T and ISO / IEC known JVT. This coding standard is also known as MPEG-4 Part10. Fig.1 gives the principle of the video compression chain H.264/AVC. In this codec, each frame will be divided into 16×16 macro blocks. Each macro block is encoded separately. In the beginning of the chain a module decision is present to decide which mode, Inter or Intra prediction, was used. Once chosen, the residual prediction error will be calculated and passed to the entire transformation. After that, all transformed coefficients will be quantified with a quantification matrix. The quantified coefficients passed to entropy coding unit using either the Context Adaptive variable Length Coding CAVLC or the Context Adaptive Binary Arithmetic Coding CABAC. The decoding chain is inserted in the encoder to reconstruct the current block or current frame used to the prediction of the next block or the next frame. So the quantized coefficients passed with the inverse quantization and the inverse transform to obtain the residual prediction error of the decoded frame. This error will be added to the prediction and the result is then passed to the deblocking filter to provide the reconstructed frame and hence the decoded video. While encrypting on H.264/AVC, we should conserve its format [2] that is to say, keep the same syntax for the bitstream. Several methods have been proposed in the literature to encrypt video using symmetric encryption algorithm such as DES (Data Encryption Standard) [3] or AES [4] (Advanced Encryption Standard). In this context the video data is protected in three ways. 14th international conference on Sciences and Techniques of Automatic control & computer engineering - STA'2013 Sousse, Tunisia, December 20-22, 2013 978-1-4799-2953-5/13/$31.00 ©2013 IEEE STA'2013-PID3 -