School of Cyberspace Security, Beijing University of Posts and Telecommunications, Beijing 100876, China

I. INTRODUCTION

With fast evolution of computing technology,more and more digital resources such as software, games, multimedia videos, films, electronic books, office documents, images and photos, and text-based resources are produced in variant fields. However the fact that the digital contents are easily being copied or shared in internet or used without licensed or authorized may do harm to the content providers or rights holder. How to protect the rights of the content producer’s, content provider’s is an important issue in the content industry, thus it need to use digital rights management(DRM)[1-4] technology to ensure all the lifecycle management such as authentication, authorization, licensing, payment, usage control, privacy protection and right violation detection and tracing[5-7], more research institutes or academy try to give a general model to solve all the problem of all kinds of digital resources, however it is impossible for different variant content formats to use one common DRM model, for example, the DRM technologies between digital images and office documents are different. In fact, unstructured data such as word processing files, design drawings is always being edited or modified, and transited frequently in local or mobile storage disks,which may be business documents, sensitive commerce content, or high-level source codes and program software, with daily work progress, meanwhile multimedia video content has its special media structure and network streaming protocol [1-3,6,8]. Upon image data, watermarks is the most popular technology can be used for copyright owner or distributor identification. Before releasing, identity or ownership information can be organized as watermark embedding in the original image data, after the image data was released to the end user, we can extract and identify the copyright or identity information[9-14].

As for the most popular reading content,e-Book has its special exchange data format(e.g. EBX, EPUB, KF8, Mobipocket, PDF,and Topaz) , how to protect the e-Book content from being restored or transferred for illegal usage is an important issue, the most security of e-Book DRM is its interoperability and adaption between different reading device[4-7,15-18]. In the Internet environment there exists much more text-based valued content, such as word, powerpoint or html pages,how to protect the content from being freely used, it needs to set secure policy and protect the content in a private network space, when the condition is satisfied, user can get the original text-based content[19], for the text-based content may be low-valued resource, thus the text-based content protection technology is usually based on resource container isolation rather than encryption[20-25].

The authors proposed a dynamic DRM model selection method for various DRM applications.

II. RELATED WORK

DRM as a technology for copy protection or copy prevention and moreover as a measures for valued digital content consumed or used in a reasonable and fair mode and can forbidden unauthorized user’s usage and consumption.Upon the DRM topic, many academy, institutes and industrial focused on much more research and proposed various solutions,productions, specifications and standards, the classifications of the DRM researches can be listed as below [6-9,11-15, 16-26].

In 2001, W3C firstly built up the group of digital rights management (DRM) as a worldwide formal organization for the digital content protection research. later ACM SIGSAC,IEEE Signal Processing sponsored special issue for DRM topic, since 2001, more than 20 international conferences focused on DRM and much more research achievements have gained [18-26].

(1) XrML Rights Expression Language Standard[19-22]

In 1994, the researcher Mark Stefik working at Xerox PARC proposed the Digital Property Rights Language (DPRL) for rights expression language , later the first edition of XrML is developed in Xerox PARC for rights expression language. In 1999, the new version XrML was improved by the new ContentGuard company founded by Microsoft and Xerox, and later when ContentGuard worked in OASIS technical committee, Electronic Book Exchange(EBX[26]), MPEG-21 standards committee,XrML was submitted and finally adopted as an official rights expression standard ISO/IEC 21000-5.

(2) OMA DRM Specification[23]

The Open Mobile Alliance sponsored mobile content protection research among worldwide mobile phone companies in 2002 , later OMA released OMA DRM specification version 1.0 and 2.0 in 2004 and 2006, in which OMA DRM 1.0 defined 3 main methods:Forward Lock, Combined Delivery , and Sep-arate Delivery, while in OMA DRM 2.0 an individual DRM Public key infrastructure (PKI)certificate is proposed for enhanced and secure mobile content consumption. OMA DRM specifications are the most important guideline for mobile content usage and limitation for mobile content, which provided a milestone of DRM research in worldwide range.

(3) Microsoft DRM[24]

Microsoft first proposed and implemented Windows DRM for WMV and WMA content protection, which implemented the DRM system by a secure container for content encryption, and the PKI certificate for license authorization. The Windows DRM SDK can allow developers integrate and apply DRM SDK to implement their own media business provider and gain business profits such as download and streaming, a business flaw of Windows DRM is that it does not support RealNetworks media format *.ram and *.rm.And later Microsoft released streaming DRM platform called Microsoft Silverlight Smooth Streaming (MSS) for online DRM service.

(4) RealNetworks Helix DRM[25]

In 2003, RealNetworks released its Helix DRM platform which supports a wide range of media formats such as RealAudio(R), RealVideo(R), MPEG-4, MP3, H.263 video, and AAC and AMR, the Helix DRM can enables PC devices, mobile devices and domain home devices. Especially Helix DRM provides native support, or transfer to secure memory for supporting multiple secure formats.

(5) Adobe Flash DRM[26]

Adobe Flash Media Server and Flash Access provide HTTP Dynamic Streaming(HDS) service as a whole solution of digital rights management (DRM) around streaming content with high level security ,the features include: 1) No exposed URLs and media file locations. 2) No client cache.3) No exposed media on the server. 4) Unique transfer protocol limits stream ripping.5) Support for SSL and encrypted streams, 6) Client authentication with Flash Media Server. The Flash media server now supporting android and iOS mobile platform.

(6) Apple HLS DRM[27]

Apple Http live Streaming DRM support an extendible and reliable streaming service protection solution which can support not only iOS devices but also support Android platform and devices, it encrypt the content payload serial pieces (TS streaming files) in tiny files format, when access the DRM-protected streaming, it need to gain the M3U8 index file and extract the DRM decrypt key to decrypt the ciphered TS steaming, the Apple HLS DRM supports PCs, iOS and Android platform and devices. Contrast to MPEG’s Dynamic Adaptive Streaming over HTTP (DASH), the HLS DRM is more available and compatible for variable mobile devices.

(7) Academic DRM Researches[28-35]

In fact, many academic DRM researchers studied variant topics, we can see the evolution of DRM technologies from Content Scrambling System (CSS), Content Encryption, Content Watermarking, Copy Prevention and High-bandwidth Digital Content Protection (HDCP)), the DRM relies on how the content encryption/scrambling/watermarking,secure authentication, authorization, privacy protection, and secure payment. H. Chang and M. J. Atallah proposed software code protection by guards in 2002, and in 2010,U. J. Jang, H. Lim, and Y. Shin proposed a license audit model for secure DRM systems in IP-based environments, H. W. Yang built up an enhanced digital rights management authentication scheme based on smart card in 2013. In 2014, Bellini et al. presented a conceptual DRM scheme in a peer-to-peer mode,which tried to integrate it with a traditional DRM system, Similarly, Das et al. presented a conceptual biometric-based DRM for mobile devices which can support anonymous and secure access, however, the 2 scheme just provided a conceptual trusted model which is not verified in real mobile platform. Shi Wenchang proposed a copy-move forgery detection approach with particle swarm optimization, in which it integrated the particle swarm optimization (PSO) algorithm into the SIFT-based framework.

III. DIGITAL RIGHTS MANAGEMENT REQUIREMENT

The main requirements of digital rights management at least include but not limited the 6 aspects, called the SACLUP DRM requirement, the SACLUP attributes includes the following 6 meanings, the DRM conception model (X+Y= Z to U) and DRM security requirements are described as Fig.1 and Fig.2:

(1) Security

(2) Authentication

(3) Constraint

(4) License

(5) Usage Control

(6) Payment

(1)Security:include content encryption,watermark, hash or digital signature of license.

(2)Authentication:content usage authentication, includes identity management by password, certificate or biometric authentication.

(3)Constraint:whether allow the content be used or not relies on the content usage conditions, for example, whether user commit valid license request data or pay the given fee, or satisfy the usage domain control or period constraint.

Fig. 1 the DRM Conception Model: X+Y= Z to U

Fig. 2 DRM security requirements

(4)License:release the valid and secure license file(such as XrML file, or authorization code) to the user who satisfied the license constraint and condition.

(5)Usage Control:accord to the license,the DRM client control the content was consumed as the license defined, violated usage may lead to the system collapse or stop to work by the DRM monitor and sensor system.

(6)Payment:the main goal of DRM is to control and reasonable benefit or money the DRM measure and tools, thus once the content itself was downloaded or released to the user, when playing, viewing or used the content , a most important constraint is to pay proper fee or the equivalent cost,such as system score or digital virtual cashes, otherwise users can use the content for playing, viewing, or reading or enter the system.

IV. DRM SECURITY INFRASTRUCTURE

In this paper, we studied various DRM model,technology and application, and first proposed DRM Security Infrastructure (DSI), in which we defined encryption, hash, signature algorithm, watermarking algorithms, authentication, usage control, trusted counter, conditional trace, secure payment. and based on the DSI we then proposed a whole classification approach and architecture of all kinds of DRMs, in which we proposed 6 typical classes of copyrights and content protection DRMs architecture: (1) Software-oriented DRM,(2) eB-ook-oriented DRM, (3) Multimedia oriented DRM, (4) Unstructured data oriented DRM,(5) Image oriented DRM, (6) Text oriented DRM. Based on the above DSI, we then proposed a dynamic DRM model selection method for various DRM application, which can be adapted dynamically for different technology of different applications, which can provide a whole solution for variant DRM development in a rapid and customized mode. The proposed DRM method, technology and application in this paper provided a common, flexible and extendable solution for variant DRM scenes,and can support rapid and customized development.

Digital rights management is a mechanism that provide reasonable and secure approach to prevent the content from being used and consumed in a control mode to ensure the content provider’s business benefit. The content or service provider first protects the raw content product to a secure status, such as encrypt or scramble the core data or embed watermark into the content, and then release the protected content or service to end user, the end user obtains the content or service, before using the protected content end user should obtain the license, the license center check whether the end user satisfies the constraint and condition,which may include obeying the usage agreement and paying for the consume of the content or service, and then the license center release personalized license which is related the user’s hardware identity environment(HID),and the license should be signed to prevent it being tampered by end user. The DRM Security Infrastructure is described as Fig.3.

V. POPULAR DRM MODEL BASED ON DSI

5.1 Software-Oriented DRM Model

(1) Software-Oriented DRM Special Requirements

Software is a special application which may require to register in the OS platform. A secure and reliable software-oriented DRM should limit the application run as end user license agreement (EULA), and can support the application migration, re-install and register according to the real requirement. Thus it should consider the fair use of the software itself, and of course there must prevent a strong mechanism from being cheated by end user.

Soft-Oriented DRM should ensure the software application or games are properly used in a valid usage, such as they can be run in a specified function status according to its license authorization rights.

Table I DRM license expression

Fig. 3 DRM security infrastructure

The license for DRM can be organized as license code, license key, license file or license USB token, and the security and performance is described as follows Table 1:

(2) Software-Oriented DRM Trusted Model

Fig. 4 Software-oriented DRM model

Fig. 5 E-Book-oriented DRM architecture

Software-oriented DRM has its special requirement then other DRM system, such as it may migrated from one computer to another, it may need to destroy the original software and prevent the owner of computer from using the software again, thus software-oriented DRM should support migration and re-installation and forbidden the original installation.

The core DRM principle of software-orient system must authenticate and authorize the software user environment and get the license dynamically according to the end user’s hardware and software unique fingerprint, such as MAC serial and hard-disk serial or CPU serial or the Hash of the three serials, thus the license is co-related with the end user,and then it can prevent the software install in any machine and identity which machine installed or to migrate the software itself. The Software-oriented DRM Model is described as Fig.4.

5.2 E-Book-oriented DRM

A most popular DRM is for internet users to read, study and utilize electronic books, digital technical reports, of which the users should obey the copyright and forbidden spreading the whole or part of the e-Book to Internet or save as to local machine in a plain mode.A fairly good e-book DRM is a trade-off between security and convenience for content provider and readers. E-book DRM is very useful for current electronic books especially in digital reading times, which can avoid paper-format books transform, delivery and release, users can easily get the book content when finish purchase.

A secure and trusted e-Book DRM should satisfy all the requirements of e-Book rights protection, such as preview, read-only, whole or part content reading, the forbidden and limitation rights and operations may include NO saving as, NO backup, NO printing, NO spreading or sharing to Internet. And the e-Book DRM should support user identity-based authentication and authorization, thus the user can read the e-book content in different devices, and once the fee is consumed out of the total payment, the system will not allow the user use the content again. A layered dis-tributed e-Book DRM model is constructed in Fig.5.

5.3 Video-oriented DRM model

The video content such as films, digital network content is very popular in current times,especially films will pay much more cost to make, to produce and manufacture, thus it needs to protect the content from being shared or transfer to internet or being recorded and redistributed for illegal business trade.

Different to other content, the video content usually be played in offline and online mode, the video content should support online streaming play with protocols such as RTP,RTSP for network users, and in some condition, it may need to support domain-license authorization for group users consumes. As for digital content protection, video protection is a very typical DRM scene among all content format. The popular online streaming DRM system for video content include Windows DRM and Silverlight, Realnetworks DRM,Adobe flash AIR, Apple HLS, which support a variant format content.

A typical trusted model of the video-oriented DRM include (1) content encryption,(2)Authentication management, (3) License management, (4) Key management, (5) Protocol supporting, (6) Pay management (especially in TV STB CAS, the standards including DVB,DVB2) . A typical video-oriented DRM trusted model is described in Fig.6 as follows:

5.4 Image-oriented DRM model

Digital image is a popular content format, its copyright and ownership and security problems become more and more serious, digital watermark is an important method for image content protection. In fact, A digital watermark is a kind technology of embedding a marker signal into a noise-tolerant carrier signal objects such as an image, audio, video data,which includes two stages: (1) Watermark embedding; (2) Watermark extracting. A typical watermark-based DRM model is described as Fig. 7.

In watermark embedding stage, the system embeds the watermark in the carrier signal under the key K with the algorithms such as LSB DCT or DWT, in which the watermark signal may be audio, pictures, video, texts or 3D model, a carrier signal can carry several different watermarks at the same time. In the watermark extracting stage, the system extracts watermark from the carrier signal source with the extracting algorithm under the key k, and according to the signal extracted from the carrier signal, the system decides whether there exists the accepted watermark or not.

The performance evaluation of a watermark system includes imperceptibility, robustness,tamper resistance, capacity. A good watermark system is a trade-off between security and effi-ciency.

Digital watermarking can be applied for ownership Identification for valued digital image or video content management, and can be used for identification management of commodity products picture in electronic commerce. And moreover digital watermarking is efficient for copyright protection for digital works such as drawings, films copyrights protection and prevention spreading without copyright holder’s permission.

Fig. 6 Video-oriented DRM architecture

Fig. 7 Image-oriented DRM Model

5.5 Unstructured data oriented DRM

Unstructured data such as word processing files, PDF files, AutoCAD design drawings or program source codes is frequently edited or modified in daily work progress, which becomes more convenient to spread and transit the intellectual property resource via online network services or copy into large capacity offline storage devices, however misused or deliberately copy of commercial entity data such as source codes, commercial plan, or other valuable and sensitive data may lead to serious business lost or data leakage, how to detect, control and protect the commercial or sensitive unstructured data is one of most important research topic of data leakage protection(called DLP).

A possible concept DLP DRM system should have the ability to prevent the content from being violated or misused by inner or outside persons in advance rather than the damage occurred and then repairing(content monitor or content filtering systems are very weak for counterfeit or cheating by inner users). An efficient and de fact secure DLP system should work in the kernel-level encryption/decryption mode, in which it should provide a thorough solution that block all possible leakage channel, monitor all usage lifecycle status and control all kinds of network and storage environment. The DLP can not only protect the sensitive data from the external intrusions, but can prevent the data producer copying or exporting sensitive data via any ways. The most basic security of the confidential unstructured data is to encrypt the data itself into cipher mode when it is produced and keep the data content in cipher status, which can only be used by authorized users, if the confidential data was illegally used,it cannot be opened, recognized, viewed, or printed, or exported. The layered architecture of kernel-level DLP system model is described in Fig. 8.

(1) Dynamic Key Management Center:to ensure the security and identity management,the DLP scheme should provide mutual authentication and dynamic key management for client and server security management. Once the client or the server is false or invalid, the DLP system can recognize the counterfeiter.

(2) Mutual Authentication Server:To ensure each both identity of the entity of the DLP client and server, the DLP system should support mutual authentication for the identity management. Thus without valid authentication, the third client or server cannot be accepted in the system, and during the mutual authentication interaction the system negotiates the session key for the data encryption.

(3) Kernel-level Mandatory Encryption Driver:The kernel-level mandatory encryption DLP should work in file system drive layer, and implement as a drive software middleware. Thus once the system is built up and works, it will work in an “encrypting while writing, decrypting when opening” mode. Of course all the operations should pass the authentication and legal authorization, otherwise the DLP middleware will not encrypt/decrypt the cipher data.

(4) Self-Protection Security Client:to ensure the DLP client drive middleware process is not killed or stopped by data producer, the DLP should provides self-protection security mechanism to enhance the client drive middleware process thus it can work in mandatory mode to avoid the risk of being bypassed the mandatory encryption policy and lead to the failure of DLP.

5.6 Web Text-oriented DRM

Text-oriented DRM its core principle is constraint all the text-based content, such as story,paper, research report into a read-only container, which can only be viewed in read-only mode, readers can NOT copy, paste, save as or print the content itself. When user want to save as or copy the content he/she should become VIP member and pay fee for the text.

The control and limitation method of the text-oriented DRM is read-only and view-only container, which is very easy to implement either in web pages or other local executable applications. Web text-oriented DRM can be viewed as a simplified-edition of e-Book-oriented DRM, such as Chinese Apabi DRM,Adobe e-Book DRM et al.. A typical text-oriented DRM secure model is described in Fig. 11 as above.

VI. FUTURE DEVELOPMENT AND DIRECTION OF DRM

6.1 Flaws of DRM itself

With the evolution and development of the human culture, content protection will rely on people’s morality, conscientiousness and law constraint rather than the technology itself ,DRM will not be the good option selection for content protection, for DRM makes it very difficult and inconvenient for people using and consuming the content, and moreover, it leads to much more complexity for service supporting and maintain, another reason is different DRM technologies are difficult in interoperation [1-5, 8-11, 16, 36-40].

Fig. 8 Unstructured data oriented DRM Architecture

Many institutes, academy researchers,computer scientists and content providers are opposed to DRM, Apple Steve Jobs and Microsoft Bill Gates criticized DRM technology in 2006 at CES, Microsoft Bill Gates pointed up that DRM causes problems for legitimate consumers while trying to distinguish between legitimate and illegitimate users. Steve Jobs announced Apple stopped requiring DRM on its iTunes Store for music content, the iTunes Store is DRM-free now for songs since January 6, 2009, which was welcomed by many customers. Current DRMs cause many problems for legitimate consumer’s usage and consumption, and most DRM only restricts consumer’s usage control but does not provide convenient migration and transfer mechanism between different devices.

Fig. 9 Web Text-oriented DRM Architecture

Fig. 10 X+ Y(without DRM) = Z to U

Fig. 11 X with Y to U

6.2 Future of DRM: DRM-free

For the DRM technology itself restricts the freedoms of people’s usage and consume of digital content, but it cannot provide convenient interoperation and application for domain or mobile consumption or usage of digital content. The future development of digital content consumption is DRM-free mode,which can support much more application scenes of content freely consumption in variant environment[41-48]. The future content consumption mode is not again X+Y=Z to U (content + DRM to Users), but with the development of human’s liberty and morality, DRM will not again will be adopted for content protection, the world will enter into a DRM-free era in the future multimedia computing time,the future content consumption is on the way with law, liberty and morality(without DRM),the future content consumption model is described in Fig.10 and Fig. 11.

VII. CONCLUSION

Upon the content consumption technology, in this paper we studied various DRM model,technology and application, and first proposed DRM Security Infrastructure (DSI), in which we defined encryption, hash, signature algorithm, watermarking algorithms, authentication, usage control, trusted counter, conditional trace, secure payment. and based on the DSI we then proposed a whole classification approach and architecture of all kinds of DRMs, in which we proposed 6 typical classes of copyrights and content protection DRMs architecture: (1) Software-oriented DRM,(2) eB-ook-oriented DRM, (3) Multimedia oriented DRM, (4) Unstructured data oriented DRM,(5) Image oriented DRM, (6) Text oriented DRM. Based on the above DSI, we then proposed a dynamic DRM model selection method for various DRM applications, which can be adapted dynamically for different technology of different applications, and can provide a whole solution for variant DRM development in a rapid and customized mode. The proposed DRM method, technology and application in this paper provided a common, flexible and extendable solution for variant DRM scenes,and can support rapid and customized development. Moreover, we proposed an opinion that the future life will enter into a new era that the content usage and consumption will not again adopt DRM technology rather than with law, liberty and morality.

[1] J. S. Erickson, “Fair use, DRM, and trusted computing”,ACM Communication, Vol. 46, no. 4, pp 34–39, April, 2003.

[2] E. W. Felten, “DRM and public policy”,ACM Communication,Vol. 48, no. 7, pp 112, July,2005.

[3] E. Becker, W. Buhse,N. Rump, “Digital Rights Management: Technological, Economic, Legal and Political Aspects”,New York, NY, USA:Springer-Verlag, 2003.

[4] D. Lindsay and S. Ricketson, ‘Copyright, privacy,and digital rights management (DRM)’, in New Dimensions in Privacy Law: International and Comparative Perspectives, A. T. Kenyon and M.Richardson, Eds. New York, NY, USA: Cambridge Univ. Press, pp 121–153, 2006.

[5] WIPO - World Intellectual Property Organization. http://www.wipo.int

[6] Digital Rights Management. http://www.w3.org/2000/12/drm-ws

[7] Digital Rights Management. http://www.acm.org/sigs/sigsac/ccs/CCS2003/drm.html.

[8] Foroughi A, Albin M, Gillard S. “Digital Rights Management: A Delicate Balance between Protection and Accessibility”,Information Science.Vol. 28, no.5, pp 389-395, 2002.

[9] B. Rosenblatt, “DRM, law and technology: An American perspective”,Information review., Vol.31, no. 1, pp 73–84, 2007.

[10] H. Chang and M. J. Atallah, “Protecting software code by guards,” ACMCCS-8Workshop on Security and Privacy in Digital Rights Management,Springer, Berlin, Germany, pp 160–175, 2002.

[11] Waller, A.O.; Jones, G.; Whitley, T., “Securing the delivery of digital content over the Internet”,Electronics & Communication Engineering Journal, Vol. 14, no.5, pp 239-248, 2002.

[12] Conrado, C.; Kamperman, F., “Privacy in an identity-based DRM system”,IEEE Proceedings of 14th International Workshop on Database and Expert Systems Applications, pp 389-395, 2003.

[13] C.H. Huang, S.C. Chuang, Y.L. Huang, J.L. Wu,“Unseen visible watermarking: a novel methodology for auxiliary information delivery via visual contents”,IEEE Transactions on Information Forensics & Security, Vol.4, no.2, pp 193–206,2009.

[14] E. W. Felten and J. A. Halderman, “Digital rights management, spyware, and security”,IEEE Security Privacy, Vol. 4, no. 1, pp 18–23, 2006.

[15] B. L. Fox and B. A. LaMacchia, “Encouraging recognition of fair uses in DRM systems”,ACM Communication,, Vol. 46, no. 4, pp 61–63, 2003.

[16] T. M. Thanh and M. Iwakiri, “A proposal of digital rights management based on incomplete cryptography using invariant Huffman code length feature,”Multimedia Systems,Vol. 20, no.2, pp 127–142, 2014.

[17] H. W. Yang, C. C. Yang, and W. Lin, “Enhanced digital rights management authentication scheme based on smart card,”IET Information Security, Vol. 7, no. 3, pp 189–194, 2013.

[18] U. J. Jang, H. Lim, and Y. Shin, “A license audit model for secure DRM systems in IP-based environments,”Journal of Information Processing Systems, Vol. 6, no. 2, pp 253–260, 2010.

[19] ODRL: Open Digital Rights Language. http://www.odrl.net

[20] InterTrust Technologies Corp. http://www.intertrust.com/main/technology/index.html.

[21] EBX, http://www.pc104.com.cn/Specification/EBX.pdf

[22] XrML Specifications. http://www.xrml.org/get_XrML.asp

[23] OMA DRM, http://www.openmobilealliance.org

[24] Microsoft Media Rights Server. Microsoft Corp,http://www.microsoft.com/windows/windowsmedia/drm/default.aspx

[25] RealNetworks, Inc, RealSystem Media Commerce Suite Technical White Paper. http://www.realnetworks.com

[26] Adobe flash access, http://www.adobe.com

[27] HTTP Live Streaming, https://developer.apple.com/streaming/

[28] Shabtai, Y.Elovici and L.Rokach, “A survey of data leakage detection and prevention solutions”,Springer Science & Business Media,2012.

[29] Sultan Alneyadi, Elankayer Sithirasenan and Vallipuram Muthukkumarasamy, “A survey on data leakage prevention systems”,Journal of Network and Computer Applications,Vol.62, pp 137–152, 2016.

[30] Sandip A. Kaiei , S.V. Kulkarni, “Data Leakage Detection”,International Journal of advanced Research in Computer and Communication Engineering, Vol.1, No. 9, pp 668-678, 2012.

[31] Kamran Morovati , Sanjay Kadam and Ali Ghorbani, “A network based document management model to prevent data extrusion”,Computers &Security, Vol.59, pp 71–91, 2016.

[32] Kwok S.H, Yang C.C, Tam K.Y, “Watermark design pattern for intellectual property protection in electronic commerce applications”,Proceedingsof the 33rd Annual Hawaii International Conference on System Sciences, 2000.

[33] Kundur, D.; Karthik, K., “Video Fingerprinting and Encryption Principles for Digital Rights Management”,Proceedings of the IEEE, Vol. 92,no.6, pp 918–932, 2004.

[34] Iwata, T.; Abe, T., “A DRM system suitable for P2P content delivery and the study on its implementation”,the 9th Asia-Pacific Conference on Communications, pp 806-811, 2003.

[35] Andreaux, J.P., Durand, A., “Copy protection system for digital home networks”,IEEE Signal Processing Magazine,Vol. 21, No.2, pp 100–108,2004.

[36] Messerges, Thomas S. and Dabbish, Ezzat A,“Digital rights management in a 3G mobile phone and beyond”,In Proc. the 2003 ACM workshop on Digital rights management, Washington, pp. 27-38, 2003.

[37] Shi Wenchang, Zhao Fei, Qin Bo and Liang Bin, “Improving Image Copy-Move Forgery Detection with Particle Swarm Optimization Techniques”,China Communications, Vol.13, no.1, pp 139-149, 2016.

[38] Reihaneh Safavi-Naini, “Import/Export in Digital Rights Management”,in Proc. the 3rd ACM workshop on Digital rights management, Washington, pp.99-110, 2004.

[39] P Koster, W Jonker, "Digital Rights Management",Springer Berlin Heidelberg, Vol 25. No.1,pp 225-235, 2007.

[40] Sultan Alneyadi, Elankayer Sithirasenan and Vallipuram Muthukkumarasamy, “A survey on data leakage prevention systems”,Journal of Network and Computer Applications, Vol.62, pp 137–152, 2016.

[41] Sandip A. Kaiei , S.V. Kulkarni, “Data Leakage Detection”,International Journal of advanced Research in Computer and Communication Engineering,Vol.1, no. 9, pp 668-678, 2012.

[42] Kamran Morovati , Sanjay Kadam and Ali Ghorbani, “A network based document management model to prevent data extrusion”,Computers &Security, Vol.59, pp 71–91, 2016.

[43] Xuyun Zhang, Chang Liu, Surya Nepal, Suraj Pandey, and Jinjun Chen, “A Privacy Leakage Upper Bound Constraint-Based Approach for Cost-Effective Privacy Preserving of Intermediate Data Sets in Cloud”,IEEE Transactions on Parallel and Distributed Systems, Vol.24, no.6,pp 1192-1202, 2013.

[45] V. Shobana and M. Shanmugasundaram, “Data Leakage Detection Using Cloud Computing”,International Journal of Emerging Technology and Advanced Engineering,Vol.3, no.1, pp 111-115, 2013.

[46] Sultan Alneyadi, Elankayer Sithirasenan and Vallipuram Muthukkumarasamy, “Detecting Data Semantic: A Data Leakage Prevention Approach”,IEEE conference on Trustcom/Big-DataSE/ISPA,Vol.1, pp 910-917, 2015.

[47] Yin Fan, Yu Rongwei and Wang Lina, “A distribution model for data leakage prevention,”International Conference on Mechatronic Sciences,Electric Engineering and Computer (MEC),pp 2617-2620, 2013.

[48] Charles Perez, Babiga Birregah and Marc Lemercier, “ The Multi-layer Imbrication for Data Leakage Prevention from Mobile Devices”,IEEE International Conference onTrust, Security and Privacy in Computing and Communications(TrustCom), pp 813-819, 2012.