Keys - Disappearing Cryptography

Cryptography Reference

In-Depth Information

the fourth position may identify Bob after it is XORed with the secret

vector.

Boneh and Shaw extend this idea by combining it with ideas from

the world of error-correcting codes. Each attacker will need to flip

many bits before the watermark is effectively erased, because the

error-correcting codes compensate for random flips.

12.7 Summary

Most algorithms in this topic can add additional security with a key.

This key can be used to create a pseudo-random bit stream by com-

puting successive values of some encryption or hash function:

f

(

key

)

,f

(

f

(

key

))

,f

(

f

(

f

(

key

)))

,....

This bit stream can be used to either choose a subset of the file or to

control how it is encoded at each location. Only someone with the

same key can recover the information. Of course, it also makes sense

to encrypt the file with another key and a cryptographically strong

algorithm before it is inserted.

Keying the algorithm defends against the often public nature of

computer standards. If the software is going to be used often and

distributed to many people, then the algorithms inside of it become

public. Adding a key to the algorithm increases its strength dramati-

cally.

The Disguise Many of the algorithms add an extra layer of disguise

by using the key to modify the behavior of the algorithm. This

means that any attacker will not be able to extract the data with

knowledge of the algorithm alone.

How Secure Is It? Pseudo-random bit streams created by repeated

encryption or hashing can be quite secure.

How to Use It Replace any random number generator used to add

noise with a pseudo-random keyed stream, which can also be

used to extract particular subsets or to rearrange the data itself.

The algorithms in Chapter 13, for instance, use a keyed encryp-

tion function to change the data's order.