Network Watermarking Location Method Based on Discrete Cosine Transform

Abstract

The idea of network digital watermarking comes from the audio digital watermarking, which plays an important role in traffic tracking. However, there are many problems in the existing watermarking schemes. For example, the selection of watermark embedding area is not combined very well with the characteristics of the data stream; the interaction between watermarks reduces the detection rate and the concealment. In this paper, we propose a method of using the Discrete Cosine Transform(DCT) domain to locate the watermarks. After the discrete cosine transform, we select the lower energy area for discrete embedding watermarks. This is a novel attempt for the DCT in network watermarking. The experimental results show that the proposed method DCTBWL (Discrete Cosine Transform Based Watermarking Locating) can effectively improve the concealment and robustness of the watermarks in network flow. Introduction With the rapidly development of internet, network security issues have become more and more seriously. Stepping stone[1], anonymous communication system and botnet[3] make it very hard to trace the attacker. In order to better solve the above problems, many scholars have put forward the active network flow watermark(ANFW) technology by drawing on the idea of digital watermarking[2]. Some special information will be hidden through changing some characteristics of the flow generated from the sender. After the network transmission, if the corresponding watermark is detected at the receiver, it is considered that there is an obvious network communication relationship between the sender and the receiver. This is a kind of active network flow shaping and analysis technology. Nowadays the embedding of watermarks always have nothing to do with the content of the packets. According to the different watermark carrier, ANFW technology mainly includes traffic rate[4], packet timing[5] and interval centroid[6~8]. However, the existing methods are not perfect, there are some problems as following: 1), the selection of watermark embedding area is not very well combined with the characteristics of the data stream. For example, the offset in [6] and [7] is only considered the time aspect and the characteristic of the flow is neglected. 2), The watermarks are embedded in a continuous area, when the front of the stream appears to be inserted or lost, all of the watermarks behind will be affected. Which result in the decrease of the watermark detection rate([5~8]). 3), It is easy for attacker to recognize the abnormity when a continuous area of data has been changed. With the low concealment, watermarks are easy to be removed or damaged. Such as the exploration for ICBW and IBW[9,10], or the exploration for DSSS-W[11,12]. Based on the above problems exist in the current watermarking schemes, in this paper, we propose a new method for network watermark location by using the discrete cosine transform(DCT). The experimental results show that this method has more robust and better concealment. The rest of this paper is organized as follows: Section 2 analyzes the feasibility of the application of DCT in network watermarking as well as the process in detail. The experimental results validating the analysis are presented in Section 4. The paper is concluded in Section 5 along with some future research directions. 3rd International Conference on Materials Engineering, Manufacturing Technology and Control (ICMEMTC 2016) © 2016. The authors Published by Atlantis Press 1515 Discrete Cosine Transform Based Watermarking Locating Scheme Feasibility Analysis of the Application of DCT in Network Watermarking. Digital watermarking technology plays an important role in the field of copyright protection and information security. The discrete cosine transform is widely used in image watermarking and audio watermarking technology[13], and the distribution of energy is presented after transform. Since it is difficult to perceive the operation in the lower energy area; the information can be hidden with high concealment. The idea of network digital watermarking comes from the image and audio watermarking technology, which is similar to a certain extent. Extracting an feature of network flow as an input signal, and then we could have the distribution of its energy by DCT. Pick up the original flow which with low energy as the watermark embedding area. Experimental results show that the discrete watermark embedding model reduces the interaction between the watermarks, and improves the robustness of the watermark in the transmission process. 0 200 400 600 80