Text Watermark Technology and Network Security Communication
He Nanzhong Li Shengtao
Huazhong University of Science and Technology Wuhan 430074, People's Republic of China
E-mail: lst_foxhust@263.net
Summary
For the rapid development of the current digital watermarking technology on the digital image watermark, this paper proposes the text watermark technology implemented by line shift coding, word movement coding, and feature encoding. This watermark technology is simple, small storage technology than the image watermark technology. The amount, high efficiency and other advantages.
Keyword
Row coding word coding feature encoding is undeniable
text
For simplicity, the model is now encoded, key, and reverse coding. That is: Information → Embedded Algorithm (Carrier) → Secure Carrier → Extraction Algorithm (Carrier) → Information.
While the electronic data exchange is exchanged, the data security requirements are increasingly required while the electronic data exchange is exchanged. We use traditional digital encryption algorithms and digital information hidden techniques to increase data security. At the same time, we use text as a carrier, which saves storage space than image watermarks while also improving operational efficiency. And in real life, the text is more frequent than image applications, more extensive. Using the text watermark not only realizes security communication, but also protects the copyright of the carrier text, and can also combine some of the traditional encryption algorithms such as digital signatures to implement digital signatures.
Algorithm implementation:
To embed the information into the text, of course, we can preprocess information, such as encryption, formatting, etc., and the imaginary information is formatted as binary code, and then embedding binary information in the text using different rules, where These rules are:
1. Run-moving coding line shift encoding, he is embedded in watermark information by vertically moving the location of the text line. Watermark information and the secret key jointly specify which text lines in the document will be moved. For example, we can specify the upload representation "0 ", Downward indicates" 1 ", of course, this is just a prior agreement. This can be embedded in one. Of course, the text line itself and the adjacent two lines should be long enough (guaranteed that it will not be indented by a small range of modifications), adjacent behavior control lines, not any movement. This can be row through the control line and mark To determine the embedded information. For example: a Thermal Lattice BGK Model with a Robust Boundary Scheme Is Developed for the BoussinesQ
INCOMPRESSIBLE FLUIDS. in the model the velocity and temperature fields are solved by TWO Independent
Lattice BGK Equations Which Are Combined Into a coupled one for the whole system. The 2D Natural
From the above information, we will apparently obtain the distance between the first two lines than the distance between the two lines. We can get the three lines of text embedded bit information as "0" according to the implementation of the agreement, and the word movement coding word. The password is embedded in the position of the characters in the text in parallel. It is similar to the row mobile coding, and we can agree on the rules of the move in advance. When using this method, the distance between adjacent characters is not necessarily the same. (Of course, we can embed the watermark information with manual methods); words move, embedding each information requires three blocks. The bitmap is a bit control bit. (Of course, this is only for the English document) If it is a Chinese document, we need to use feature encoding. For example: Convection Flow Of Air In a Differential Heated CAVITY OF Aspect Ratio 4 IS Simulated for Values of the Rayleigh Number Up to 1010.
During the simple period, we did not embed watermarks with the interval of password directly. If the text is left to "1", the right shift to "0", we can know that the information bits of the above text embedded are "0110100" 2, and the feature encoding feature encoding is embedded by changing the various properties in the text. The feature can be the location of the character, the font, paragraphs, character height, and the like (in fact, the above-mentioned row coding and word shift encoding are feature coding categories). We can also replace some vocabulary in text through synonyms; The spacing of the vocabulary can be defined to control the embedding of information. For example: Huazhong University of Science and Technology
The above information Watermark We use the character size to embed, of course, we have more ways. The characters are larger than the left side, as "1", vice versa "0"; so that we only need two characters to embed a message. We only embed information in the first text Huazhong, apparent that the first character h is a small number of words, and the second character is four words. The third character is four words, the fourth character is a small four word ... Our watermark information is obviously "1001".
For easy understanding, the paragraph and character spacing have changed significantly. If in regular text, this difference is not easy to see. The above information is embedded in us, we don't need the original document to extract watermark information, calling blind hidden. We are of course not embedded from the first or first character to the last or one character without the rules, we must use the secret key K to determine the location or method of the information embedded. Such information may hide each place in the text, thereby improving security. Document Watermark Information Security:
To make the document watermark information, it must be done in conjunction with the complete cryptographic system and transport protocol. We must ensure the integrity of watermark information, prevent third-party interception, modification, deletion, and reconstruction. Since the watermark information and document information are inseparable, I am in the protection of watermark information to the protection of documents.
Document data security:
The security of document data can be divided into document security and network communication on a single machine. The single-machine document security can use the classic encryption algorithm to ensure that many other articles mentioned, we focus on Talk about the document security on the network communication
1. Assumption: There is a third party, he can intercept, modify, delete and reconstruct all information transmitted on the network.
The security of information is only based on the security of the private key, ie, the encryption algorithm and the communication protocol are disclosed, and the encryption algorithm is very strong.
The above two assumptions are very strong. If secure communication can be achieved on the two assumptions, the security communication on the network is also possible.
2, realize
The definition of the document integrity of network communication is, ie the sender sends a document. After the recipient receives the document, it can determine the source, content and whether the content and whether or not it is replaced, modified. In communication, we also use the common key encryption system. Ensure that the integrity of communications can be attributed to preventing intermediaries initiated by third parties on the network.
Intermediary attack: (send party: Alice Receiver: Bob Third Party: Mallory)
(1) Alice transmits her public key to Bob. Mallory intercepts this key and transmits its public key to Bob. (2) Bob transmits his public key to Alice. Mallory intercepts this key and transmits his public key to Alice. (3) When Alice will send the secret to Bob with "Bob" public key, Mallory intercepts it. Since the information is actually encrypted with the public key of Mallory, he decrypts its own private key. Re-encrypt the information again with the public key of BOB and transfer it to Bob. (4) When Bob will transfer it to Alice with the "Alice" public key encrypted, Mallory intercepts it. Since the information is actually encrypted with his own public key, he decrypts the information with his private key, and then re-encrypts the public key of Alice and transfer it to Alice. This attack is also possible even if the public key of Alice and Bob is stored in the database. Mallory can intercept Alice's database query and use its own public key instead of BOB's public key, and he can do the same thing for Bob, with its own public key instead of Alice's public key. He can also secretly invade the database, replace Alice and Bob keys with his own key. Next, he simply waited for Alice and Bob to talk to each other, then intercept and modify the information, he succeeded! Let's take a look at the chain agreement:
(1) Alice transmits her public key to Bob. (2) Bob transmits his public key to Alice. (3) Alice encrypts her packet with Bob's public key and transmits half of the encrypted message to Bob. (4) Bob encrypts his packets with the public key of Alice, and transmits half of the encrypted packets to Alice. (5) Alice will send the other half of the packet to BOB. (6) Bob combines Alice's two semi-packets and decrypts his private key; BOB transmits him to the other half of the packet to Alice. (7) Alice combines BOB two-half message and decrypts her private key. The important point here is that only half of the packets, no other half, the packet is useless. BOB can only read the Alice packet (6), and Alice can only read the BOB packets only by step (7). There are many ways to implement it: (1) If a packet encryption algorithm is employed, half of each group (for example, every one bit) can be sent in every half message. (2) The decryption of the packet depends on the initial vector, and the initial vector can be sent in the other half of the packet. (3) The first half of the packets that may be transmitted may be the one-way HASH function of the encrypted message, and the encrypted message itself may be the other half. Although the chain protocol can prevent third parties from re-sending the information. But you can't prevent him from reading. According to two assumptions, the third party inevitably knows that if you want to implement secure communication, only the re-communications receiver and sender have a shared key or information (directly with shared key decryption It is possible to send it); but this is a conflict, but in real life, we still have a way to do, we can distribute such as letters, such as letters, etc. Under these two assumptions, we cannot achieve security communication, but we can increase the difficulty of third-party attacks. This is what we can do.
We can do this with multiple communication protocols and secret segmentation agreements. (1) Alice divides her public key through the secret segmentation protocol into N group. Bob only receives all of the N groups or some of them to restore the ALICE's public key. (2) Alice sends all key packets to BOB through multiple communication protocols. That is, through different channels, for example, the packet information is sent to BOB; (3) Bob, the same method is transmitted to Alice. (4) Alice encrypts her packets with Bob's public key, and divides the secret segmentation protocol into M group. (5) Alice also transmits the packet 1 of the encrypted packet to BOB with multiple communication protocols. (6) Bob encrypts her packets with Alice public key, and divides the secret division protocol into M group. (7) Bob also transmits the packet 1 of the encrypted packet to Alice with multiple communication protocols. (8) Alice transmits other packets to BOB, respectively. (9) Bob transmits other packets to Alice. ......... (6) Bob Received the last group to combine all packets of Alice, and decrypt him; Bob will send him the last set of packets to Alice. (7) Alice combines all the members of BOB and decrypts her private key. If you want to bring more trouble for Mallory, we can randomly select communication mode in the channel collection, and send a message secret packet to send a secret split and multiplexed protocol.
Some questions discussions:
Text Watermark Theory and image watermarks, but lacks classic processing methods such as Xiang Fu Liye transform, so that in some respects, it is lacking than the digital image. For example, in anti-interference. But we can use a combination of multiple embedded methods to effectively solve this problem. Of course, the text handling method of the image processing method should we deserve further discussion and research.
Reference Bibliographic Information Hidden Technology Implementation and Application Application Computer Cryptography
