Cryptography Reference
In-Depth Information
Both the sender and the receiver of a message have to carry this individual key
with them.
This takes us straight to
problem 2
: handling the key. How does the key get
to the receiver? How should it be stored? This is not practical in most cases.
If you want to use a one-time pad to encrypt a complete hard disk, you need
a second disk only to store the key. How about locking away the critical disk
in the first place? Anyway, fast data streams over lengthy periods cannot be
encrypted in this way.
The following historical episode should teach us a lesson in this respect: one-
time pads were used by Soviet spies in the USA during World War II. They
reused the same pads; the KGB must have supplied them in identical batches.
That was a fateful mistake. Of course, the adversary listened in on them and
copied the messages, although they couldn't initially make sense of them. Using
simple statistical tests, they managed in arrears to fish out usable ciphertext
pairs. If you use different pads,
S
1
and
S
2
, for two plaintexts,
P
1
and
P
2
, then
the results,
C
1
=
S
2
, are independent, and both the sum
and the difference of
C
1
and
C
2
produce equally distributed random numbers.
In the case of
S
1
=
P
1
+
S
1
and
C
2
=
P
2
+
S
2
, however,
C
2
-C
1
=P
2
-P
1
holds, and the difference of two texts in the same language is by no means
random — it has striking statistical characteristics. One single little piece of
plaintext from a message will allow you to apply the zigzag method, using all
messages encrypted with the same pad,
S
1
!
This was how the USA started getting an idea of the entire dimension of Soviet
espionage in the USA for the first time after the end of World War II (rumors
had it that there were about 200 spies). The cryptanalysis was extraordinarily
difficult, because the messages had first been encrypted by means of code
books, and not all cover names were revealed. Only part of the ciphertexts
had been decrypted. Eventually the NSA gave up on their work in 1980 (!).
Still, this enterprise, named the
VENONA Project
, was a success. Famous
personalities like Julius and Ethel Rosenberg and the nuke spy Klaus Fuchs
fell victims to the project.
This example helps us understand why national intelligence organizations listen
in on encrypted texts even when they initially can't make sense of them. And it
shows very impressively that even the single most secure ciphering method is
