Multimedia Encryption

Introduction

Multimedia technology becomes more and more popular in today’s digitized and networked world. Many multimedia-based services, such as pay-TV, remote video conferencing, medical imaging, and archiving of government documents, require reliable storage of digital multimedia files and secure transmission of multimedia streams. In addition, in course of the recent booming of diverse multimedia functions/services provided by consumer electronic devices and digital content providers, more and more personal data are created, transmitted, and stored in multimedia formats, which also incur increasing concerns about personal privacy (i.e., multimedia data security). To fulfill such an overwhelming demand, encryption algorithms have to be employed to secure multimedia data.

Apart from concerns about data security, there also exist serious concerns about copyright protection issues, which are mainly raised by multimedia content providers as a hope to protect their multimedia products or services from pirate copies and unauthorized distributions. Digital watermarking is the main technique to realize such a function, by embedding digital patterns in multimedia products to be detected.

Multimedia encryption and digital watermarking constitute the kernel of digital rights management (DRM) systems. Recently, a lot of efforts have been made to define DRM systems for multimedia encoding standards. Two ISO/IEC standards have officially been released in the past three years: JPSEC (Security Part of JPEG2000) in 2004 and MPEG-4 intellectual property management and protection (IPMPX, eXtensions) in 2006. To ensure flexibility and renewability, both standards define only a framework and interfaces between different modules so that any available tool can be freely chosen by the content providers/owners in a real implementation. In this way, a malfunctioning encryption or watermarking component can be replaced by a new one without changing other parts of a system.

In recent years, some surveys have been published about multimedia encryption (Furht & Kirovski, 2004; Furht, Muharemagic, & Socek, 2005; Uhl & Pommer, 2005; Zeng, Yu, & Lin, 2006). In this article, we will also introduce some very new results that are not covered in previous surveys.

Multimedia Encryption: Why?

Modern cryptography has been well developed since 1970s. A large number of ciphers have been proposed, among which some have been standardized and widely adopted all over the world, that is, Data Encryption Standard (DES), Advanced Encryption Standard (AES) and Rivest-Shamir-Adleman public-key encryption algorithm (RSA). So, it seems natural to use any established cipher to encrypt a multimedia file/stream bit by bit. This simple and easy approach is called naïve encryption in the literature and has been used in some DRM systems. However, naïve encryption does not suit many multimedia-related applications, due to some special features required in these applications.

The first problem is that many traditional ciphers cannot run fast enough to fulfill the needs of real-time multimedia applications. For example, for video-on-demand (VoD) services, generally there are always a large number of videos stored in many servers, and a large number of video streams transmitted from these servers to end users. In this case, the encryption loads of the VoD servers may be too high to ensure smooth running of the services. Another scenario is about medical imaging systems, in which lossless compression algorithms (instead of lossy algorithms) may have to be used due to legal considerations. This means that the compression efficiency will be limited, so the resulting multimedia data will be much more bulky and the encryption load will be much higher. In applications of this kind, total (full) encryption of multimedia data (i.e., naïve encryption in term of the amount of encrypted data) should be avoided and selective (partial) encryption is suggested.