Master, most valuable exam deck

0.0(0)
Studied by 0 people
call kaiCall Kai
learnLearn
examPractice Test
spaced repetitionSpaced Repetition
heart puzzleMatch
flashcardsFlashcards
GameKnowt Play
Card Sorting

1/100

encourage image

There's no tags or description

Looks like no tags are added yet.

101 Terms

1
New cards

DES facts?

64-bit block, 56-bit key, 16 rounds.

2
New cards

AES facts?

128-bit block; 128/192/256-bit keys; 10/12/14 rounds.

3
New cards

Feistel equations?

L_i=R_(i-1); R_i=L_(i-1) XOR f(R_(i-1),k_i).

4
New cards

Feistel decryption?

Same structure, round keys reversed.

5
New cards

Confusion mainly comes from?

Nonlinear substitution / S-boxes.

6
New cards

AES diffusion mainly comes from?

ShiftRows and MixColumns.

7
New cards

AES round order?

SubBytes -> ShiftRows -> MixColumns -> AddRoundKey.

8
New cards

Final AES round omits?

MixColumns.

9
New cards

Why is DES weak?

56-bit key is too short.

10
New cards

Double DES attack?

Meet-in-the-middle.

11
New cards

ECB weakness?

Equal plaintext blocks give equal ciphertext blocks; patterns leak.

12
New cards

CBC core idea?

XOR plaintext with previous ciphertext before encryption; IV for first block.

13
New cards

OFB vs CFB?

OFB feeds cipher output and is synchronous; CFB feeds ciphertext and is self-synchronising.

14
New cards

CTR advantage?

Counter-based, hides repeats, parallelisable.

15
New cards

Preimage resistance?

Given digest D, infeasible to find M with H(M)=D.

16
New cards

Second-preimage resistance?

Given M, infeasible to find M != M' with the same hash.

17
New cards

Collision resistance?

Infeasible to find any distinct M,M' with H(M)=H(M').

18
New cards

Generic collision work for an n-bit digest?

About 2^(n/2).

19
New cards

256-bit digest generic collision security?

About 2^128 work.

20
New cards

Why must hash collisions exist?

Hashing is many-to-one: arbitrary-length inputs map to fixed-length outputs.

21
New cards

Merkle-Damgard h0?

A fixed IV.

22
New cards

Merkle-Damgard recurrence?

h_i = H(h_(i-1), M_i).

23
New cards

Merkle-Damgard final digest?

h(M)=h_n.

24
New cards

What happens to the final block in Merkle-Damgard?

It is padded and the message length is appended.

25
New cards

MAC vs MDC?

MAC uses a secret key; MDC is an unkeyed check/hash.

26
New cards

CBC-MAC core idea?

CBC-process the message and use the final output block as the MAC.

27
New cards

Why is H(k||m) risky with Merkle-Damgard?

Length extension: the digest exposes the final chaining state.

28
New cards

Password storage with salt?

Store

29
New cards

Is a salt secret?

No; it is a known random value stored with the digest.

30
New cards

Why use salts?

They defeat/prevent reuse of precomputed digest tables across passwords.

31
New cards

Hash-based PRNG recurrence?

S0=H(S), then S_i=H(S_(i-1)); extract a small number of bits.

32
New cards

Three PRG properties?

Randomness, unpredictability, unreproducibility.

33
New cards

Blum Blum Shub update?

X_(n+1)=X_n^2 mod m.

34
New cards

What secures Blum Blum Shub?

Difficulty of factoring m=pq.

35
New cards

Why not use BBS as a fast cipher?

It is too slow.

36
New cards

When does b have an inverse mod n?

Exactly when gcd(b,n)=1.

37
New cards

How do you calculate a/b mod n?

a*b^(-1) mod n.

38
New cards

What does XGCD give?

x,y with gcd(a,b)=xa+yb.

39
New cards

How does XGCD give a^(-1) mod n?

If 1=xa+yn, then x mod n is a^(-1).

40
New cards

What is phi(n)?

The number of residues modulo n relatively prime to n; |Z_n*|.

41
New cards

phi(p) for prime p?

p-1.

42
New cards

phi(pq) for distinct primes p,q?

(p-1)(q-1).

43
New cards

General totient formula?

If n=product p_i^e_i, phi(n)=product p_i^(e_i-1)(p_i-1).

44
New cards

Euler's theorem?

If gcd(a,n)=1, then a^phi(n)=1 mod n.

45
New cards

How does Euler help with modular powers?

Reduce the exponent modulo phi(n), when the base is invertible.

46
New cards

Fermat's Little Theorem?

If p is prime and p does not divide a, then a^(p-1)=1 mod p.

47
New cards

Purpose of square-and-multiply?

Efficient modular exponentiation using the binary exponent.

48
New cards

CRT condition for unique reconstruction?

The moduli must be pairwise coprime.

49
New cards

CRT core idea?

Compute modulo smaller coprime factors, then reconstruct the unique result.

50
New cards

Quadratic residue q mod n?

Some x exists with x^2=q mod n.

51
New cards

If n=pq and a is a QR in Z_n*, how many roots?

Exactly four.

52
New cards

How do you find roots mod pq?

Find roots mod p and mod q, then combine sign choices with CRT.

53
New cards

Exam trigger: a/b mod n?

Find b^(-1) with XGCD, then multiply by a.

54
New cards

Exam trigger: huge exponent mod n?

Look for phi/Euler or Fermat reduction, then square-and-multiply.

55
New cards

Exam trigger: all roots mod pq?

Factor n, solve mod each prime, combine four sign pairs with CRT.

56
New cards

RSA modulus?

n=pq for large primes p,q.

57
New cards

RSA phi(n)?

(p-1)(q-1).

58
New cards

Condition on RSA e?

gcd(e,phi(n))=1.

59
New cards

RSA private exponent d?

d=e^(-1) mod phi(n).

60
New cards

RSA public key?

(e,n).

61
New cards

RSA encryption?

c=m^e mod n.

62
New cards

RSA decryption?

m=c^d mod n.

63
New cards

RSA problem?

Given c,n,e with c=m^e mod n, recover m.

64
New cards

Why is RSAP no harder than factorisation?

Factor n, compute phi(n), find d, then m=c^d mod n.

65
New cards

Efficient RSA exponentiation algorithm?

Square-and-multiply.

66
New cards

How does CRT speed RSA private operations?

Exponentiate mod p and q separately, then recombine with CRT.

67
New cards

RSA signature idea?

Hash the message and apply the private-key operation to the digest.

68
New cards

RSA signature verification?

Apply the public-key operation to the signature and compare with the message hash.

69
New cards

What do digital signatures provide?

Authentication and non-repudiation.

70
New cards

Do signatures provide confidentiality?

No.

71
New cards

Why sign a hash instead of the whole message?

The digest is smaller, so signing is more efficient.

72
New cards

Toy RSA private-key question: first steps?

Factor n, compute phi(n), then d=e^(-1) mod phi(n).

73
New cards

Toy RSA encryption/decryption: method?

Write the modular exponentiation and use square-and-multiply; use CRT if helpful.

74
New cards

Two main security-protocol goals?

Entity authentication and key exchange.

75
New cards

How should a protocol exam answer be structured?

Objectives, exact protocol steps, security analysis.

76
New cards

What does a nonce provide?

Freshness / replay protection.

77
New cards

What does a timestamp provide?

Time-based freshness, usually requiring synchronised clocks.

78
New cards

NSPK objective?

Mutual authentication using public keys.

79
New cards

NSPK steps?

1 A->B:{N_A,A}_K_B+; 2 B->A:{N_A,N_B}_K_A+; 3 A->B:{N_B}_K_B+.

80
New cards

What breaks original NSPK?

Lowe's interleaving attack.

81
New cards

Diffie-Hellman objective?

Agree a shared symmetric key over an insecure channel.

82
New cards

DH messages?

A sends g^x mod p; B sends g^y mod p.

83
New cards

DH shared key?

g^(xy) mod p.

84
New cards

Main weakness of unauthenticated DH?

Man-in-the-middle attack.

85
New cards

NSSK objective?

Mutual authentication plus a trusted-server-generated session key K_AB.

86
New cards

NSSK steps?

1 A->T:A,B,N_A; 2 T->A:{N_A,B,K_AB,{K_AB,A}_KBT}_KAT; 3 A->B:{K_AB,A}_KBT; 4 B->A:{N_B}_KAB; 5 A->B:{N_B-1}_KAB.

87
New cards

NSSK freshness weakness?

An old compromised session key/ticket can be replayed because B cannot directly prove freshness.

88
New cards

What is Kerberos based on?

Needham-Schroeder Secret-Key.

89
New cards

Main Kerberos difference from NSSK?

Timestamps and explicit session-key/ticket lifetimes.

90
New cards

Kerberos trusted server name?

Key Distribution Centre (KDC).

91
New cards

Kerberos steps?

1 A->T:A,B,N_A; 2 T->A:{K_AB,A,L}_KBT,{K_AB,N_A,L,B}_KAT; 3 A->B:{K_AB,A,L}_KBT,{A,T_A}_KAB; 4 B->A:{T_A}_KAB.

92
New cards

What is L in Kerberos?

Session-key/ticket validity period.

93
New cards

Does Kerberos rely on timestamps?

Yes.

94
New cards

Otway-Rees objective?

Mutual authentication plus trusted-server generation of a fresh session key.

95
New cards

Does Otway-Rees need synchronised clocks?

No; it uses nonces for freshness.

96
New cards

Otway-Rees steps?

1 A->B:N,A,B,{N_A,N,A,B}_KAT; 2 B->T:N,A,B,{N_A,N,A,B}_KAT,{N_B,N,A,B}_KBT; 3 T->B:N,{N_A,K_AB}_KAT,{N_B,K_AB}_KBT; 4 B->A:N,{N_A,K_AB}_KAT.

97
New cards

What is N in Otway-Rees?

A transaction identifier binding A and B data to the same run.

98
New cards

What do N_A/N_B do in Otway-Rees?

Guarantee freshness of the server response/session key.

99
New cards

Otway-Rees weakness in the notes?

Type-flaw attack.

100
New cards

What is a type-flaw attack?

Data of one type is misinterpreted as another type.