| home / multimedia / video / 1 |    [previous][next] |
|
|
|
In the previous screenplay, you learned about some of the pitfalls of DRM. At least you know that the hacks are out there, but surely some of the DRM solutions out there on the marketplace offer the right level of security, don't they? To answer that question, let's understand conceptually what DRM solutions need to offer.
All DRM systems for digital media have several basic elements:
Figure 7-7 illustrates part of a media supply chain and the various devices that have to cooperate.
Note that DRM does not necessarily imply a seller and buyer; it merely involves enforcing rules about digital objects. But it turns out that "merely enforcing rules about digital objects" is easier said than done. A DRM system can work only when all, not some, devices in a content delivery chain enforce the rules. This
is why DRM is such a difficult proposition--it requires the cooperation of a multitude of devices, the so-called "media value chain," any one of which can potentially break the entire security of the system.
Figure 7-7: The basic elements of a DRM system.
One way to understand the challenges is to look at where DRM has failed. MP3 is a fantastically successful file format that has evaded DRM protection. Why? Consider these facts about MP3:
To make an effective DRM system, then, we must:
Does this sound difficult? This evolution is in progress right now. And in some ways, it's not new. In other parts of the computer industry, consortia of computer vendors have traditionally worked together to enforce copy protection, and in the hardware world, this was easy to enforce. Macrovision has made VCR-to-VCR copying difficult for years; double-deck VCRs didn't arrive until it no longer mattered.
Whereas the entertainment industry has deep ties and control over consumer electronics for media, the computer industry has not been so agreeable to adding expensive and mandatory copy protection to every computer and computer-connected device in existence.
Case in point: DVD burners. A DVD is just a high-capacity format with a single compelling use: storing large amounts of data on a single disk. But the fact that the movie industry has adopted DVDs as the platform of choice for consumer sales and that there are compelling non-infringing consumer uses of the format (long-playing music collections, hard disk backup, home movie distribution) means that DVD-based copy protection is not adequate to prevent the (legal or illegal) copying of copyright content, as shown in Figure 7-8.
Figure 7-8: DVDs were secure until personal computers and DVD burners entered the picture.
MPEG-21 is an MPEG sub-group that is working on standardizing rights definition and enforcement for MPEG media. The challenge faced by the MPEG-21 group is not so much a technical one but a political one: Getting a majority of vendors to agree upon a common rights definition language and enforcement system, and then ensuring that it is adopted.
Any DRM system must include these high-level pieces. After we achieve an ideal arrangement of well-expressed rules and dutifully obedient devices, however, we must still deal with a technical problem. One critical part of the media value chain--end users--might try to work around the system. This puts DRM in the interesting position of trying to securely deliver data to users on the one hand, and trying to stop these users from breaking the security and subverting the rules on the other.
Note: MPEG-21 is not a successor to MPEG-4; MPEG-21 is the numerical designation of the groups. There is also an MPEG-7 group that is standardizing content descriptions for searching and copyright. The in-order numbering scheme seems to have ended.
Microsoft's website contains a page with this innocuous sounding sentence:
When a consumer acquires an encrypted digital media file from a website, he or she must also acquire a license that contains a key to unlock the file before the content can be played (http://www. microsoft.com/windows/windowsmedia/howto/articles/drmarchitecture.aspx).
We spend much of the rest of this chapter exploring what is meant by the concepts in this sentence.
In most scenarios, when you download a movie you are entering into a limited contract, much as when you buy a ticket to a movie theater. That ticket grants you the right to enter the theater and watch the movie during a specified showing--not all day long, and not with a video camera in your jacket. It's a contract, and it's enforced by the ushers who can kick you out if they notice you sneaking into a movie you didn't pay for. Likewise, downloading a movie is a contract granting you the right to watch the movie under certain restrictions. In Internet movie viewing, the usher is DRM, but of course the usher can't see what you're doing at home, which makes the job of policing user behavior particularly difficult.
So what are the rules? In the theater, as we've said, you watch one movie at a specified time, you can't videotape the movie while you're sitting there, and you have to buy the theater's absurd prices for popcorn and soda. Here are some possible rules for online viewing:
So by paying money, or otherwise agreeing to some terms, "the consumer acquires an encrypted media file" (we talk about the nature of this encryption in the next section) and a "license that contains a key to unlock the file." The license contains the rules that the user previously agreed to. As for that key, we discuss keys in depth in the section, "Tools in the Encryption Toolbox."
The whole essence of DRM is contained in these two concepts: the encrypted media file and the key that unlocks the file (see Figure 7-9). Underlying these simple concepts is some fairly sophisticated computer science. You can't really evaluate DRM claims or have confidence in the security of your content unless you understand how encryption is applied.
Figure 7-9: Encryption.
Encryption is the world of spies and secret messages, intelligence, and counter-intelligence. Although DRM's realm of content delivery and secure transactions is more prosaic, it does have a cloak-and-dagger heritage. Encryption, in its many forms, can be described simply as the locking of data with a password. Encryption makes data unreadable until it is decrypted or unlocked with a digital password called a key.
The point of locking the data is to make the message secure. Headquarters wants to be sure that an enemy who intercepts a message sent to an agent won't be able to read it. So the message is encrypted with a special code, or key, which lets the agent unlock the message but not the enemy. If the enemy wants to read the message, they need to steal the key from the agent or analyze the message well enough that they can figure out the key.
Thieves generally do not "break" locks. They pick them, or they go through a window. The same analogy holds for encryption. Breaking an encryption system altogether would be to find some weakness that renders it weak or useless for all systems that use it. For widely trusted encryption systems (used by finance, the military, and so on), this is unlikely; the systems have been designed and attacked by the equivalent of master locksmiths for decades and have withstood the attacks.
What can be broken, though, are short passwords. They are much more vulnerable to being picked because they can be quickly guessed. Computers can try every conceivable combination of bits to guess the password very quickly. A password that is 4, 8, or 16 bits in length is far too short to provide security. What is needed are very long keys used in such a way that it takes a computer some time to try each one. If it takes a computer only a fraction of a second to make one guess at a key, but there are trillions of possible keys, it still will take perhaps hundreds or thousands of years for that computer to try all the possibilities.
Encryption technology gets weaker as computers become faster, but not that much weaker. Older encryption standards lasted for decades before it was even possible to build a machine that could guess all the keys in a lifetime. Now, modern computers can crack some of the oldest standards, and even newer standards with short key lengths are weak enough to cause concern. Thus, better encryption standards and longer keys continue to be developed.
Two measurements involved in assessing encryption vulnerability are time to crack, and cost to crack it. If, 10 years ago, it took a $10 million machine to crack a key in less than a few hundred years, that same key might be crackable by a $10,000 machine today. The number of people who could afford to crack your key has risen exponentially; so has the risk of compromise.
| home / multimedia / video / 1 | [previous][next] |
The Network for Technology Professionals
Legal Notices, Licensing, Permissions, Privacy Policy.
Advertise | Newsletters | E-mail Offers
|
Video: How System Center Helps Reduce IT Costs IBM Cloud Computing for Developers Virtual Event, April 6-8 |
Microsoft Video: OCS and Exchange: Platform Futures MORE WEBCASTS, PODCASTS, AND VIDEOS |
Created: March 27, 2003
Revised: May 24, 2004
URL: http://webreference.com/programming/xsltweb2/1
