1 00:00:01,020 --> 00:00:11,820 Let's try to enhance the security of D.S. to do this you can for example use a variant called D.S. x. 2 00:00:11,860 --> 00:00:15,640 The idea is to keep the algorithm while fixing some major problems. 3 00:00:17,040 --> 00:00:22,140 Making kingsize longer means that the number of operations required to test all possible keys will rise 4 00:00:22,140 --> 00:00:23,340 exponentially. 5 00:00:27,270 --> 00:00:31,400 The Morath operation of the DE ESX algorithm is as follows. 6 00:00:32,700 --> 00:00:38,360 The first block of input is Exel armed with the first 64 bit key. 7 00:00:38,390 --> 00:00:43,580 The result is encrypted using a 56 bit key. 8 00:00:43,600 --> 00:00:45,760 The result is then X ored with the third key. 9 00:00:45,760 --> 00:00:51,620 That also has 64 bits. 10 00:00:51,740 --> 00:00:54,090 In theory having three keys. 11 00:00:54,140 --> 00:01:02,180 Their summary length is 184 bits the size however does not contribute directly to the ciphertext security 12 00:01:03,920 --> 00:01:10,070 cryptographic analysis shows that the practical security the key is entropy boosted from 64 to more 13 00:01:10,070 --> 00:01:11,690 or less 88 bits. 14 00:01:13,190 --> 00:01:14,720 This is a large step forward 15 00:01:19,080 --> 00:01:28,030 ESX is considerably more secure than D.S. the algorithm is relatively fast which is why it was implemented 16 00:01:28,030 --> 00:01:29,830 in the EFI system. 17 00:01:29,870 --> 00:01:32,820 A file encryption system used in earlier Windows systems 18 00:01:38,030 --> 00:01:40,700 and other variant of D.S. is yes. 19 00:01:40,900 --> 00:01:42,170 Triple D s. 20 00:01:43,990 --> 00:01:45,290 The name says it all. 21 00:01:46,210 --> 00:01:51,810 The D.S.O. algorithm is applied in the solution three times to each data block. 22 00:01:51,820 --> 00:01:53,520 The mechanism is as follows. 23 00:01:56,210 --> 00:01:58,200 First a block is encrypted like in D. 24 00:01:58,210 --> 00:02:06,330 Yes it is then decrypted using a different key and finally the result is encrypted using yet another 25 00:02:06,330 --> 00:02:10,200 key as you remember. 26 00:02:10,200 --> 00:02:19,000 Encryption is symmetric and data encryption standard this operation doesn't really tighten the security. 27 00:02:19,040 --> 00:02:30,340 The second phase of the operation does not affect or improve the overall security of the ciphertext. 28 00:02:30,340 --> 00:02:36,750 The main downside of using Triple D Yes is the Scifres low speed. 29 00:02:36,830 --> 00:02:40,270 It's possibly the slowest symmetric algorithm in use today. 30 00:02:41,250 --> 00:02:44,840 All other surfers are much faster and most of them are more secure. 31 00:02:48,620 --> 00:02:52,890 The practical security of trippled Yes is about one hundred bits of strength. 32 00:02:56,630 --> 00:02:59,990 There are also other symmetric ciphers. 33 00:03:00,120 --> 00:03:06,230 Many of them belong to the RC run Rivest Code family. 34 00:03:06,410 --> 00:03:12,870 If you're guessing that the ciphers were created by Ron Rivest turns out you're right. 35 00:03:13,090 --> 00:03:20,850 The first one most showcases are C-2 this special purpose cipher was developed in 1987 for the software 36 00:03:20,850 --> 00:03:24,210 company Lotus. 37 00:03:24,350 --> 00:03:29,390 It was kept secret for some time at first until a researcher once again proved the principles we discussed 38 00:03:29,390 --> 00:03:38,380 earlier are true and published the ciphers mode of operation on a usenet group. 39 00:03:38,560 --> 00:03:48,160 RC to supposedly strong security was debunked the implemented block size with 64 bits it's too short. 40 00:03:49,880 --> 00:03:54,550 The cipher is also susceptible to chosen plaintext attacks that we talked about earlier. 41 00:03:56,860 --> 00:04:00,750 Let's now comment on the attack categories. 42 00:04:00,910 --> 00:04:07,120 The first basic type of cryptanalysis attack relies on an attacker having access to the ciphertext only 43 00:04:08,440 --> 00:04:17,740 this type of attack is completely passive an attacker has access to the ciphertext it must crack it. 44 00:04:17,760 --> 00:04:23,160 The second class of attacks known as plaintext attacks involves the assumption that an attacker has 45 00:04:23,160 --> 00:04:28,450 access both to the plaintext and to the ciphertext. 46 00:04:28,460 --> 00:04:30,590 This is actually not infrequent at all. 47 00:04:31,890 --> 00:04:37,380 If you encrypt something in a network an attacker can in most cases rightly assume that some information 48 00:04:37,380 --> 00:04:38,430 is heavily patterned 49 00:04:41,060 --> 00:04:48,050 an e-mail message very often starts off with high or when you encrypt data using SSL and IP protocol 50 00:04:48,050 --> 00:04:53,550 header containing your PC's IP address will be at the start. 51 00:04:53,590 --> 00:05:01,740 In most cases an attacker has in fact access to at least part of the plaintext and the ciphertext. 52 00:05:01,910 --> 00:05:07,910 The third attack category plaintext injection involves a manipulation of the plain text in order to 53 00:05:07,910 --> 00:05:09,880 observe resulting ciphertext. 54 00:05:11,540 --> 00:05:16,370 An attacker may send a crafted message to you without raising suspicions to make you encrypt it and 55 00:05:16,370 --> 00:05:17,460 send it back. 56 00:05:18,880 --> 00:05:20,820 This happens quite often as well. 57 00:05:23,220 --> 00:05:27,250 It's crucial for the system to be immune against the two letter attack models. 58 00:05:29,840 --> 00:05:34,520 All adequate algorithms are now secured against the ciphertext only cryptanalysis attack 59 00:05:37,540 --> 00:05:42,520 the true security of the RC algorithm is estimated at thirty four bits. 60 00:05:42,520 --> 00:05:44,550 This means that it can be cracked instantly 61 00:05:51,950 --> 00:05:53,460 another version of the algorithm. 62 00:05:53,480 --> 00:05:57,320 RC 5 was developed in 1994 for RSA 63 00:06:01,540 --> 00:06:09,290 The algorithm is very flexible many variables used in the scheme can be altered. 64 00:06:09,420 --> 00:06:15,310 You can for example specify a round number block size in Keeling's. 65 00:06:15,540 --> 00:06:22,120 There can be from one to two hundred and fifty five rounds and block size can be from 32 to 128 bits. 66 00:06:23,680 --> 00:06:27,820 Keys can even be up to two kilobytes. 67 00:06:27,830 --> 00:06:30,910 The mechanism used in RC 5 is depicted in the slide 68 00:06:36,380 --> 00:06:39,170 will now discuss the algorithms mode of operation briefly 69 00:06:42,550 --> 00:06:43,020 first. 70 00:06:43,030 --> 00:06:45,250 A key has to be divided into subkey. 71 00:06:47,040 --> 00:06:50,580 The number of subcase must be equal to the selected number of rounds. 72 00:06:53,560 --> 00:06:59,130 Then as usual split the message block into two even pieces. 73 00:06:59,260 --> 00:07:03,390 The two halves are combined with the first two subcase the round keys. 74 00:07:05,710 --> 00:07:13,850 Then perform the exo are and transpose bits bits reordered in the two halves. 75 00:07:14,080 --> 00:07:20,800 RC 5 in its variant called RC 6 were entered into a contest to choose the D.S. algorithm's replacements 76 00:07:20,800 --> 00:07:28,980 Sipher the did not come close to winning both algorithms are assessed negatively and because of this 77 00:07:29,190 --> 00:07:31,690 they are not considered particularly secure.