IT 366: Lecture 3

How many keys are in the DES keyspace?

(2^56), or 72 million billion.

How many key bits are actually used? Why?

56 bits (7 bytes) of the 64-bit key are used, with the remaining 8 bits used for parity.

What is a mode of operation?

A means of using a cipher algorithm.

How should a key be communicated by the sender to the recipient? How must it not be?

A secure channel must be used to send the key to the recipient - not transmitted 'in the clear' - anyone who learns the key value can decrypt any ciphertext ever encrypted with the key.

What are some important difference between DES and AES?

All operations are byte-oriented in AES, DES is bit-oriented The plaintext and ciphertext block size in AES is 16 bytes (=128 bits) A block is treated as a 4x4 matrix of bytes The operations performed in each 'round' are: A table-lookup substitution A row-based transposition A column-based substitution A bitwise-XOR substitution, using the subkey for that round. AES is not a Feistel cipher - there are separate encryption and decryption processes.

How does CBC mode work? What should it be used for?

Cipher Block Chaining: The IV is XOR'ed with the plaintext, creating ciphertext. In the second cycle, the ciphertext is XOR'ed with the second block of plaintext, creating more ciphertext. The result of each XOR is also encrypted with a function; even if the plaintext is the same for each block. the output is still pseudorandom. This continues for all blocks - CBC is commonly used to provide confidentiality, can also be used to generate a MAC (Message Authentication Code).

How does CFB mode work? What should it be used for?

Cipher Feedback Mode (Streaming Mode 1): An IV is used to initialize the shift register; the ciphertext from the previous cycle is fed into the right end of a shift register, and the contents of this register are encrypted, and 's' bits of the result are XOR'ed with the plaintext to produce the ciphertext. CFB is used to provide confidentiality in a streaming environment.

How does Counter mode work? What should it be used for?

Counter Mode: An initial value of the 'counter' needs to be shared confidentially between the encrypter and decrypter. The counter is encrypted (in both cases), and then XOR'ed as we saw with CFB and OFB. In each subsequent cycle, the counter value is changed to a new value, most commonly be incrementing it by 1, modulo the block size. Because each cycle is independent of any other, multiple cycles can be performed in parallel for high-speed calculations. The encryption algorithm is used to pseudo-randomize an incrementing series of values. These results are used to pseudo-randomize the plaintext block values. Counter Mode is used to provide confidentiality, and for software/hardware efficiency.

What type of cipher is DES? What is its block size? Key size?

DES (Data Encryption Standard) is a symmetric block cipher. It is a block cipher because it processes a block of plaintext with every cycle; DES encrypts plaintext in blocks of 64 bits, to produce ciphertext blocks of the same size, using 56 blocks of a 64-bit key. If the plaintext elements are bytes (text, one 64-bit block is equivalent to 8 bytes, and the 56-bit key is equivalent to 7 bytes. )

DES is no longer an approved Federal standard - why should we study it?

DES is the best-known symmetric encryption scheme.

What is Double DES? Does it reduce to a single stage? What type of attack is feasible against it?

Double DES encryption of plaintext: C = E(K2, E(K1, P)), P = D(K1, D(K2, C)) Does not reduce to a single stage - produces a different set of mappings between plaintext block values and ciphertext block values, and is probable much more secure than single DES. However, if we know or suspect a related pair of plaintext and ciphertext values, we can attempt a 'meet in the middle' attack by encrypting the plaintext value with every possible key, decrypting the ciphertext value with every possible key, and testing for a matching result - we would then know the key values.

How does ECB mode work? What should it be used for?

Electronic Codebook: Lookup plaintext value in a codebook; book gives ciphertext. ECB is only used with very short messages and unpredictable plaintext.

Why has brute force search time declined over time?

Hardware advancement: Moore's Law (density of transistors doubles every 18 months) - DES can be cracked in a matter of hours.

What is the impact if a key is "cracked"? How can this damage be limited?

If a key is 'cracked', an opponent can decrypt any ciphertext ever encrypted with that key. This damage can be limited by changing key values often, limiting the effects of a 'broken' key. Be aware that information needs to be protected until it no longer has value, and even very strong encryption may not be adequate.

How many keys need to be tested in an "average" brute force attack?

On average, half of the keyspace must be tested.

What is a challenge when using symmetric encryption?

Sharing the key in a secure manner can be difficult.

Why is symmetric encryption called "symmetric"?

Symmetric Encryption is so called because decryption is essentially the symmetric opposite of encryption. It uses the inverse operations, in the reverse order, with the same key, as used in encryption.

What is the Feistel design? What are some characteristics of a Feistel cipher?

The Feistel Design is a product cipher - a scheme that combines substitution and permutation, usually in multiple iterations ('rounds') of a sequence. A number of identical rounds transform the data from the previous round (or the original plaintext). Properties: The same algorithm is used for both encryption and decryption. This is possible because the algorithm is 'self-symmetric'. The advantage is that only one algorithm needs to be developed, not two. Each round uses a sub-key, derived by permutation of the key. The complexity of the sub-key derivation adds to the security of the scheme. The processing in one round is relatively simple, but the large number of repetitions of the round logic adds to the security of the scheme (DES uses 16 rounds).

How does confidentiality of the shared key affect confidentiality of the plaintext?

The security of a symmetric cipher scheme is predicated on the confidentiality of the key - anyone who learns the key value can decrypt any ciphertext ever encrypted with that key.

Why not use the "secure channel" in the formal model for sending plaintext?

This is because secure channels are either too cumbersome, expensive, or too small of bandwidth to adequately transmit messages (plaintext).

What is the TDEA? Why is the middle step the opposite of the outside two steps?

Triple Data Encryption Algorithm, or 3DES The middle step is opposite of the two outside steps in order to improve the security of encrypted data, while leveraging our investment in DES - allowing us to convert from single DES to 3DES gradually, rather than requiring an instantaneous 'hot cut'.

Should the symmetric algorithm chosen be kept secret? Why / why not?

Typically, knowledge of the algorithm used for encryption is not secured - it can be public. The security of a symmetric cipher scheme is not significantly enhanced by concealing the algorithm details (and it makes it hard to manage if you try). The security of a symmetric cipher scheme is predicated on the confidentiality of the key - anyone who learns the key value can decrypt any ciphertext ever encrypted with that key.

What key/s is/are used in symmetric encryption?

Wherever the shared key was used in encryption, it is used in decryption to 'undo' the encryption operation - symmetric encryption requires the encrypter and decrypter to use the same key; sharing the key in a secure manner can be difficult.

A Feistel cipher is symmetric - how can it use an operation that is not reversible?

The inverse of bitwise XOR is bitwise XOR - the inverse of a swapping operation is a swapping operation, and the same function is used for both encryption and decryption.