A more secure SEAL was designed in 1992 by
Rogaway and Coppersmith, speci cally for the purpose of obtaining a
software effi cient stream cipher. SEAL is still the fastest steam cipher
for software implementations on contemporary PC's, with "C"
implementations running at 5 cycle/byte on common PC's (and 3.5 cycle/byte on
some RISC workstations).

The design of SEAL shares many similarities with the design of common block ciphers. It is built around a repeating round function, which provides the "cryptographic strength" of the cipher.

Algorithm optimized for 32-bit processors.

- Jogging requires eight 32-bit registers and a cache a few
kilobytes
- Through a seemingly slow operation, SEAL preprocesses the operation
of keys in a set of tables.
- Then these tables are used to increase the speed of encryption and
description

SEAL is not a traditional flow cryptosystem, this is a pseudo functions family, is a family of functions pseudo -
random.

**ITS OPERATION**

Its operation is based on a initial process in
which are calculated values for some tables from the key, so that proper
encryption can be carried out in a way really fast. Unfortunately, it is also
subject to patent algorithm.

Given a key 160-bit, k and a number n of 32-bit

- SEAL extends a string of L n k bits (n)
- L can take any value less than 64 kilobytes
- Is supposed to SEAL takes advantage of the property of k is selected randomly
- Then k (n) may be that not distinguished
computationally a random function of L 'n' bits

A very useful feature of this algorithm is that it
is not based on a linear system of

generation, but that of a family of pseudorandom
functions, in such a way that any part of the sequence can be calculated
providing only a whole number n of 32-bit.

With a cryptosystem as SEAL

- You can access any bits of the generated key

- Make a hard drive
- Encrypt by and every sectors of 512 bytes
- With SEAL is possible to encrypt the contents of the sector n, with an XOR of the content with the key k (n).

SEAL is based on the use of the SHA algorithm (see
section 13.1.4 on) to generate the tables

that it uses internally. In fact, there are two
versions of the algorithm, the 1.0 and 2.0, that differ precisely in that the
first employs SHA and the second its on revised version, SHA-1.

Algorithm controlled by three tables derived keys:
R, y, S, and T

- Preprocessing maps the key k, to these values
using a procedure based on SHA

Also used four 32-bit registers: A, B, C and D

- Initial values are determined by n and R and T tables
- These records are modified through several iterations, each one
involves 8 turns iterations, each one involves 8 laps.
- In each iteration 9 bits of a register are used to index the table
T.
- Content table gives a xor with the contents of a second record A,
B, C and D.
- First record is travel (shifted) 9 positions.

In some iterations the second record is modified
amended through an xor with the first now travel record

- After 8 iterations of this, A, B, C and D are added to the stream
of the key
- The iteration is completed by adding to A and C additional values
depending on n n1 n2 additional depending on n, n1, n2, n3 and n4 values.
- Exactly which depends on the parity of the number of iteration.

Function Seal

Referencia:

En lo de ejemplo y vulnerabilidades pudieras haber mejorado. Van 5 pts.

ResponderEliminar