4/29 - Maybe computers can never be 100% secure with reasonable certainty. There is always a possibility that an attacker knows something that the writers of software do not. It is also possible that an exploit is discovered and published prematurely before the software manufacturer has a chance to respond and develop a patch—so-called zero-day exploits.

Computer encryption, however, is different. With encryption, one's data either is safe, or it's not. Encryption is all, or it's nothing. There is very little middle ground where something is just less secure or more secure.

Encryption, done right, is so strong that it can protect content from the most powerful governments in the world. It is "security that may protect against military agencies for years to come" (Bruce Schneier, Applied Cryptography, p. xix).

Encryption also protects the intellectual property of entertainment companies (IP in this case stands for intellectual property, not Internet Protocol). Not all DVDs are encrypted, but most are. DVDs will eventually be superseded by the new Blu-ray and HD DVD technologies. Blu-ray and HD DVD, too, are protected by a cryptographic system.

The encryption that protects movie DVDs was broken a long time ago. Now, if encryption can offer airtight protection, and if the encryption that protects movie DVDs was quickly broken, one may well ask, what happened? Moreover, the ease with which the initial DVD encryption was defeated begs the question, does the same fate await Blu-ray discs and HD DVDs?

First a few words on the basics of encryption and cryptography.

Encryption takes a mathematical function and applies it to content in such a way that the the content is re-arranged and transposed to where, in the case of entertainment discs, it cannot be played. A mathematical function in this context is a "cryptographic algorithm, also called a cipher" (Schneier, p. 2).

Encrypted content is thus scrambled. In fact, the name of the scheme that protects encrypted DVD movies is Content Scrambling System (CSS).

In order to play an encrypted DVD, the DVD has to be decrypted. In decryption, a related mathematical function is applied to the content that restores the content to its original, unencrypted form.

Here's where it gets interesting. One's first instinct in safeguarding content might be to keep the cryptographic algorithm a secret, so that no one discovers the algorithm and decrypts the content. That might be someone's initial impulse. However, the attempt to safeguard data by hiding the way in which an algorithm works is a case of what is known as security through obscurity, and security through obscurity is almost never an effective solution. "If the strength of your new cryptosystem relies on the fact that the attacker does not know the algorithm's inner workings, you're sunk" (Schneier, p. 7).

So, in order to come up with a secure cryptographic algorithm, one must do the counter-intuitive thing, and put the algorithm out there for the whole world to see. That way, the cryptographic community can scrutinize the algorithm and either make suggestions to make the algorithm secure or discard it altogether, if it is seriously flawed.

It is therefore one of the first principles of cryptography that the algorithm be made known. One assumes that an attacker has complete knowledge of the way in which an algorithm works. Consequently there is another first principle in cryptography that says that all of the security in an algorithm "is based in the key" (Schneier, p. 3).

A key is simply one of many possible values. An algorithm uses a key to encrypt, and a related algorithm uses a key to decrypt. The keys may be the same, or they may be different, depending on the system. If the keys are the same, this is known as a symmetric algorithm. PKI (Public Key Infrastructure) crypto, however, uses a different key to encrypt than is used to decrypt. The systems that protect DVDs, Blu-ray, & HD DVD use both symmetric and public key algorithms.

The objective of an attack in cryptanalysis is to restore the original unencrypted contents, without prior "access to the key" (Schneier, p. 5).

If the number of possible values for your key is very small, then your system is at risk, and it can be broken simply by trying every possible value of the key until the attacker happens upon the right value. This is known as a brute-force attack.

If, on the other hand, the number of possible values is very large, then it may take an attacker a very long time before he or she decrypts the content by trying every possible value. And I mean a very long time—as in the next ice age, the sun going nova, or the time it will take for all matter to collapse into black holes (examples are taken from Schneier's Applied Cryptography, p. 18). If it takes that long, your shouldn't have to worry about your data.

The actual systems used to protect DVDs, Blu-ray, & HD DVD are much more complex than a simple encryption / decryption scenario, however. In PKI, you encrypt with someone else's public key and that someone decrypts using his or her very own private key. If you encrypt with your private key, it does not make your content secure—anyone can decrypt with your public key—but it does prove that the message came from you and no one else, something known as authentication. The systems used by entertainment discs make use of all of these elements—symmetric keys, PKI, authentication—and more.

One thing encryption is not is analog protection. Once a DVD is decrypted, the content is output and sent to speakers and displays, along either analog or digital paths. If a digital path, the content can be protected once again by encryption. If analog streams are used to output the contents, however, encryption cannot be used, and the system has to resort to other non-cryptographic measures for protection. Computer encryption can only deal with the digital stream.

Analog protection is the province of Macrovision and other companies, which do a good job of protecting content on the analog side of the equation. Analog protection is most important to any overall system for protecting the contents of DVD, Blu-ray, or HD DVD discs. However, it is outside the scope of this writing. Ths essay only deals with encryption. To analog protection, it has nothing to say. It's out-of-scope.

Another thing encryption is not is a tripping mechanism. Several tools can decrypt DVDs. Techniques are also used to make these tools fail. This is protection, but it is not encryption. To these technologies, the current writing has nothing to say. It's not encryption.

Given the principles of cryptography, there is no reason why the digital protection of entertainment discs cannot be made airtight. The system that protects both Blu-ray and HD DVD appears to be an attempt to do just that. Even so, truth is stranger than fiction. If nothing else, it should be interesting to watch how the story unfolds in the coming years.

Even though encryption either is very secure or it's not, there are nevertheless some unknown variables that preclude mathematical certainty. An organization that devises a cryptographic algorithm could place a back door in the algorithm. There is always the chance of advances in cryptanalysis. Breakthrough's in computing could make attacks that are computationally infeasible today doable tomorrow.

Computer encryption deals mainly with computational security, rather than unconditional security (Schneier, p. 8). The theoretical possibility of decrypting by brute-force forces encryption into the realm of the computationally infeasible, rather than the unconditionally secure, given infinite resources. Even so, one can only conclude that content is very secure if, in order to break the encryption, it takes "a billion times the age of the universe" (Schneier, p. 9).

To Be Continued ...