Stream Ciphers
#1

[attachment=12325]
Stream Ciphers
• Generate a pseudo-random key stream &
xor to the plaintext.
• Key: The seed of the PRNG
• Traditional PRNGs (e.g. those used for simulations) are not secure.
E.g., the linear congruential generator:
Xi = a Xi-1 + b mod m
for some fixed a, b, m.
It passes the randomness tests, but it is predictible.
Linear Feedback Shift Registers
Feedback shift register:
(“register”, “feedback”, “shift”)
LFSR: Feedback fnc. is linear over Z2 (i.e., an xor):
Very compact & efficient in hardware.
• Stream Ciphers from LFSRs
Desirable properties of f:
– high non-linearity
– long “cycle period” (~2n1+n2+...+nk)
– low correlation with the input bits
• Example LFSR-Based Ciphers
• Geffe Generator:
– Three LFSRs
– LFSR1 is used to choose between LFSR2 & LFSR3:
y = (x(1) Ù x(2)) Å (Øx(1) Ù x(3))
– Correlation problem: P(y = x(2)) = 0.75 (or, P(y = x(3)))
• Stop-and-Go Generators:
– One (or more) LFSR is used to clock the others
– E.g.: The alternating stop-and-go generator:
Three LFSRs. If x(1) is 0, LFSR2 is forwarded; otherwise LFSR3. Output is x(2) Å x(3).
LFSR-Based Ciphers (cont’d)
• The Shrinking Generator:
– Two LFSRs
– If x(1) is 1, output x(2).
Else, discard both x(1) & x(2); forward the LFSRs.
• A5 (the GSM standard):
– Three LFSRs; 64 bits in total.
– Designed secretly. Leaked in 1994.
– A5/2 is completely broken. (Barkan et al., 2003)
• E0 (Bluetooth’s standard encryption)
– Four LFSRs; 128 bits in total.
GSM A5/1
• The A5/1 stream cipher uses three LFSRs.
• A register is clocked if its clocking bit (orange) agrees with one or both of the clocking bits of the other two registers. (majority match)
• Software-Oriented Stream Ciphers
LFSRs slow in software
• Alternatives:
– Block ciphers (or hash functions) in
CFB, OFB, CTR modes.
– Stream ciphers designed for software:
• RC4
• SEAL
• RC4
(Rivest, 1987)
• Simple, byte-oriented, fast in s/w.
• Popular: MS-Windows, Netscape, Apple,
Oracle Secure SQL, WEP, etc.
Algorithm:
• Works on n-bit words. (typically, n = 8)
• State of the cipher: A permutation of {0,1,...,N-1}, where N = 2n, stored at S[0,1,...,N-1].
• Key schedule: Expands the key (40-256 bits) into the initial state table S.
• RC4 (cont’d)
The encryption (i.e., the PRNG) algorithm:
i ← 0
j ← 0
loop: {
i ← i + 1
j ← j + S[i]
S[i] ↔ S[j]
output S[S[i] + S[j]]
}
• SEAL
Software-Optimized Encryption Algorithm
• Rogaway & Coppersmith, 1992, IBM
• Extremely fast in software
Speed comparisons:
(from Crypto++ 5.1 benchmarks, on a 2.1 GHz P4):
• RC4 & WEP
WEP: Wired Eqv. Privacy (802.11 encryption prot.)
• RC4 encryption, with 40–104 bit keys.
• 24-bit IV is prepended to the key; RC4(IV || k). IV is changed for each packet.
• Integrity protection: By encrypted CRC-32 checksum.
(What are some obvious problems so far?)
• Key management not specified. (Typically, a key is shared among an AP and all its clients.)
• Design process: Not closed-door, not very public either.
Attacks on WEP
(Borisov, Goldberg, Wagner, 2000)
Obvious problems:
• 24-bit IV too shot; recycles easily. (And in most systems, implemented as a counter starting from 0.)
• CRC is linear; not secure against modifications.
• Even worse: Using CRC with a stream cipher.
Passive decryption attacks:
• Statistical frequency analysis can discover the plaintexts encrypted with the same IV.
• An insider can get the key stream for a packet he sent (i.e., by xoring plaintext and ciphertext); hence can decrypt anyone’s packet encrypted with the same IV.
Authentication: challenge-response with RC4
• server sends 128-bit challenge
• client encrypts with RC4 and returns
• server decrypts and compares
• Problem: attacker sees both the challenge & the response; can easily obtain a valid key stream & use it to respond to future challenges.
An active attack:
• Since RC4 is a stream cipher, an attacker can modify the plaintext bits over the ciphertext and fix the CRC checksum accordingly.
• Parts of the plaintext is predictable (e.g., the upper-layer protocol headers).
• Attacker sniffs a packet and changes its IP address to his machine from the ciphertext.
(If the attacker’s machine is outside the firewall, the TCP port number could also be changed, to 80 for example, which most firewalls would not block.)
• Hence, the attacker obtains the decrypted text without breaking the encryption.
A table-based attack:
• An insider generates a packet for each IV.
• Extracts the key stream by xoring the ciphertext with the plaintext.
• Stores all the key streams in a table indexed by the IV. (Requires ~15GB in total.)
• Now he can decrypt any packet sent to that AP.
Note: All these attacks are practical. Some assume a shared key, which is realistic.
• The final nail in the coffin:
(Fluhrer, Mantin, Shamir, 2001)
The way RC4 is used in WEP can be broken completely: When IV is known, it is possible to get k in RC4(IV || k).
• WEP2 proposal: 128-bit key, 128-bit IV.
This can be broken even faster!
Replacements for WEP
• WPA (inc. TKIP)
– encryption: RC4, but with a complex IV-key mixing
– integrity: cryptographic checksum (by lightweight Michael algorithm)
– replay protection: 48-bit seq.no.; also used as IV
– WPA2 (long-term replacement, 802.11i std.)
– encryption: AES-CTR mode
– integrity: AES-CBC-MAC
Reply

Important Note..!

If you are not satisfied with above reply ,..Please

ASK HERE

So that we will collect data for you and will made reply to the request....OR try below "QUICK REPLY" box to add a reply to this page
Popular Searches: characteristics of advance block ciphers, what is sa 285 cfb, an improved symmetric key cryptography with dna based strong ciphers, kiturami ctr 5500, stream ciphers algoritm for matlab, ctr 5500, stream ciphers and banking,

[-]
Quick Reply
Message
Type your reply to this message here.

Image Verification
Please enter the text contained within the image into the text box below it. This process is used to prevent automated spam bots.
Image Verification
(case insensitive)

Possibly Related Threads...
Thread Author Replies Views Last Post
  STREAM COMPUTING computer science crazy 2 1,621 19-04-2011, 12:16 PM
Last Post: seminar project
  Aurora: A New Model And Architecture For Data Stream Management summer project pal 1 1,104 23-01-2011, 09:19 PM
Last Post: summer project pal
  STREAM COMPUTING A SEMINAR REPORT Computer Science Clay 2 3,370 27-11-2010, 01:02 PM
Last Post: projectsofme
  Stream computing seminar projects crazy 0 2,684 13-06-2009, 03:18 PM
Last Post: seminar projects crazy

Forum Jump: