1 00:00:01,550 --> 00:00:04,170 Let's now give several examples. 2 00:00:04,370 --> 00:00:06,350 Encryption is not a new technique. 3 00:00:07,660 --> 00:00:14,090 From the beginning of history people tried to hide and encrypt all sorts of information a good example 4 00:00:14,090 --> 00:00:22,240 of this is Caesar cypher which will be brought up on more than one occasion later Caesar cypher which 5 00:00:22,240 --> 00:00:28,120 according to historians was really used by Julius Caesar was a code designed to hide an intended message. 6 00:00:28,210 --> 00:00:36,040 For example battlefield commands in a ciphered way messages would be encrypted using a simple technique 7 00:00:37,640 --> 00:00:40,300 letters would be shifted by some number of positions. 8 00:00:40,370 --> 00:00:48,680 For example three positions if a shift of three was used d would be written instead of a the algorithm's 9 00:00:48,690 --> 00:00:56,080 key and Caesar cipher is the shift number this encryption method is extremely weak for several reasons. 10 00:00:58,910 --> 00:01:00,860 Above all the key is very short. 11 00:01:01,890 --> 00:01:08,760 To test all possibilities it's enough to check 26 Schiff's the number of letters in the alphabet when 12 00:01:08,760 --> 00:01:10,280 running permutations. 13 00:01:10,320 --> 00:01:13,660 You'll need to see if the result forms an intelligible message. 14 00:01:18,960 --> 00:01:25,830 And apart from this the ciphertext would reveal much about the plaintext some plaintext data will always 15 00:01:25,830 --> 00:01:27,000 be revealed. 16 00:01:27,030 --> 00:01:28,740 Length is a good example. 17 00:01:29,770 --> 00:01:34,320 The longer the ciphertext the longer the plaintext is. 18 00:01:34,340 --> 00:01:37,540 This can be assumed on a one to one relationship. 19 00:01:37,670 --> 00:01:40,480 It's practically linear. 20 00:01:40,530 --> 00:01:42,480 This is inevitable to a degree. 21 00:01:43,930 --> 00:01:49,770 If you don't want this information to transpire you'd always need to generate maximum length ciphertext 22 00:01:52,140 --> 00:01:59,870 this solution is implemented in cryptography practically only by three letter agencies in the US. 23 00:02:00,010 --> 00:02:05,600 The machines in these agencies are hooked by a specially dedicated link that is used for incessant pseudorandom 24 00:02:05,770 --> 00:02:09,010 data transmission. 25 00:02:09,080 --> 00:02:13,970 If you tried to listen on this link you'd not be able to discern neither the beginning nor the end of 26 00:02:13,970 --> 00:02:19,600 an intended message in the noise that complements transmission. 27 00:02:19,610 --> 00:02:26,560 The point is that Caesar cipher reveals also other information letter frequency distribution is a trait 28 00:02:26,560 --> 00:02:32,050 shared by all languages some letters are more or less common. 29 00:02:32,070 --> 00:02:38,410 It's easy to check this if you use an encryption method similar to Caesar cipher. 30 00:02:38,410 --> 00:02:44,550 Some characters in the ciphertext will come up more frequently using this assumption. 31 00:02:44,560 --> 00:02:52,290 It's really easy to deduce that plain text representation of the most frequent ciphertext letter scissors 32 00:02:52,290 --> 00:02:55,290 cipher was an example of a weak encryption technique. 33 00:03:00,450 --> 00:03:03,990 At the other end there is an all out ideal solution. 34 00:03:03,990 --> 00:03:10,970 One time pad it's ideal on several counts. 35 00:03:10,980 --> 00:03:19,430 First of all it's impossible to crack the encryption makes use of a one time key pad. 36 00:03:21,120 --> 00:03:23,120 A key is never reused. 37 00:03:24,790 --> 00:03:29,560 This is the first of the requirements that are hard to insure because the difficulty connected with 38 00:03:29,560 --> 00:03:34,230 computer generation of pseudo random numbers. 39 00:03:34,310 --> 00:03:36,700 The second requirement is even harder to me. 40 00:03:37,540 --> 00:03:41,690 The key must be the same length as the plain text. 41 00:03:41,720 --> 00:03:48,500 If this condition is satisfied if you have a 200 character message in a 200 character random key encrypting 42 00:03:48,500 --> 00:03:52,180 the message is trivially easy. 43 00:03:52,380 --> 00:04:00,030 It's enough to perform an exo or operation exclusive disjunction with individual key and message bits. 44 00:04:01,960 --> 00:04:08,860 The result of the X or operation is one only if the values of two parameters keep it Message bit are 45 00:04:08,890 --> 00:04:09,560 different. 46 00:04:11,800 --> 00:04:17,350 If the bits are the same the result is zero something is altered. 47 00:04:17,420 --> 00:04:21,220 But this can be reversed at this point. 48 00:04:21,220 --> 00:04:28,190 This is vital for us a message that is encrypted using one time pad doesn't reveal any information about 49 00:04:28,190 --> 00:04:32,500 the key or the plain text message. 50 00:04:32,530 --> 00:04:39,140 So why isn't one time pad encryption and universal use. 51 00:04:39,220 --> 00:04:42,870 This is because the Scifres applicability is very limited. 52 00:04:43,000 --> 00:04:50,200 They depend on the possession of a pseudo random key even if it appears otherwise at first glance. 53 00:04:50,200 --> 00:04:51,930 All computers use today. 54 00:04:52,240 --> 00:04:56,160 Quantum computers are not yet available are always deterministic. 55 00:04:58,330 --> 00:05:02,290 They don't have access to any random operations. 56 00:05:02,490 --> 00:05:09,010 Even if it doesn't seem so for a user they always go through the same operations in a given situation. 57 00:05:10,420 --> 00:05:18,480 So how can you generate or ensure the required random information. 58 00:05:18,480 --> 00:05:23,660 This is a serious issue that can be solved in a range of solutions. 59 00:05:23,670 --> 00:05:28,850 The first concept involved being brought from the outside from physical phenomena. 60 00:05:29,070 --> 00:05:34,650 If a computer would be for example able to measure outside temperature with high accuracy count the 61 00:05:34,650 --> 00:05:38,640 flies that sit atop a building and divide one by another. 62 00:05:38,640 --> 00:05:43,880 The result would truly be random not influenced by any other factor. 63 00:05:45,200 --> 00:05:52,640 But computers aren't capable of doing that random number generation could also be provided from keystroke 64 00:05:52,640 --> 00:05:58,720 timing dissolution was widespread but quite weak. 65 00:05:58,800 --> 00:06:05,790 The intervals between individual keystrokes from a good unsystematic typist are deterministic pace variation 66 00:06:05,790 --> 00:06:12,830 is in the order of milliseconds in advance typing keystrokes dynamics always have the same flow and 67 00:06:12,830 --> 00:06:13,430 speed 68 00:06:16,240 --> 00:06:23,990 and other sources mouse cursor coordinates a generator can use a saved coordinate assuming that a user 69 00:06:23,980 --> 00:06:30,800 has operations very the solution truly provides relative randomness. 70 00:06:30,870 --> 00:06:37,850 But what if you need a random key directly at system startup and operating system needs to generate 71 00:06:37,850 --> 00:06:45,810 a 128 bit key before a user move the mouse a bit or submitted enough keyboard input. 72 00:06:45,830 --> 00:06:53,000 This is alleviated by hardware solutions some chips can measure physical parameters with high accuracy 73 00:06:53,450 --> 00:07:00,830 temperature electromagnetic radiation signal rate and so on and use the values to generate random numbers. 74 00:07:03,020 --> 00:07:10,140 This type of hardware isn't popular yet so there are still problems with random number generating If 75 00:07:10,140 --> 00:07:16,380 a system has collected some amount of pseudo random complicated information they have to be distributed 76 00:07:16,380 --> 00:07:18,240 and managed economically. 77 00:07:19,770 --> 00:07:23,910 That can't be distributed out for the first long key. 78 00:07:23,920 --> 00:07:27,050 There's also another problem with encryption. 79 00:07:27,150 --> 00:07:32,220 If a recipient has to decode an encrypted message they have to be provided with a key. 80 00:07:32,310 --> 00:07:34,030 There's no other way around it. 81 00:07:36,270 --> 00:07:43,710 The message itself is 100 percent secure how to forward a kid to a recipient in a safe manner use a 82 00:07:43,710 --> 00:07:51,300 trusted channel key length is equal to plain text length if you have found a trusted way of exchanging 83 00:07:51,300 --> 00:07:59,190 a key way encrypted in the first place you could have directly transmitted it over the secure channel 84 00:08:00,880 --> 00:08:01,840 One-Time Pad. 85 00:08:01,910 --> 00:08:08,920 Although sophisticated is impractical let's try to strike a good balance between Caesar cipher and one 86 00:08:08,920 --> 00:08:17,630 time pad the next parts of the lecture will be dedicated to this the ideal will be the same each time 87 00:08:19,740 --> 00:08:25,860 a key has to be significantly shorter than the information that is protected using the key. 88 00:08:25,870 --> 00:08:31,570 The assumption is that it's less demanding to protect an exchange a relatively short key than to protect 89 00:08:31,570 --> 00:08:40,100 an exchange of potentially long message think of protecting a car preventing the theft of a car if it's 90 00:08:40,100 --> 00:08:43,130 otherwise unprotected might be quite difficult. 91 00:08:44,550 --> 00:08:50,160 It's much easier to protect the car with a key that you take with you everywhere you assume that if 92 00:08:50,160 --> 00:08:54,120 you have the key the car that it is used to protect is also secure. 93 00:09:00,960 --> 00:09:05,820 The module will talk about the characteristics of symmetric encryption and discuss the flaws and strengths 94 00:09:05,820 --> 00:09:06,960 of this solution. 95 00:09:08,570 --> 00:09:14,030 Will review block cipher modes or in other words tell you why even the most secure symmetric ciphers 96 00:09:14,030 --> 00:09:15,830 will fail if ill and lamented 97 00:09:20,150 --> 00:09:27,220 an alternative to block ciphers stream ciphers will also be included in the guide will touch upon stream 98 00:09:27,220 --> 00:09:34,300 ciphers in security and tell you why the number of their applications is thinning. 99 00:09:34,310 --> 00:09:40,610 Another problem covered will be public key cryptography or asymmetric cryptography a cryptographic field 100 00:09:40,610 --> 00:09:48,190 with a much shorter history while symmetric encryption was used back in the days of Julius Caesar public 101 00:09:48,190 --> 00:09:50,570 key encryption has emerged quite recently. 102 00:09:50,800 --> 00:09:53,080 It's about 100 years old. 103 00:09:53,080 --> 00:09:59,570 The practical applications of this solution are even newer. 104 00:09:59,620 --> 00:10:04,600 We'll also discuss why asymmetric ciphers aren't good for encrypting long messages and broached the 105 00:10:04,600 --> 00:10:07,330 main shortcomings of asymmetric ciphers 106 00:10:10,670 --> 00:10:16,160 another topic will include hashing functions and their mode of operation and tell you why calculating 107 00:10:16,160 --> 00:10:23,350 a hash function doesn't ensure a vital data aspect known as non repudiation not a repudiation was mentioned 108 00:10:23,380 --> 00:10:26,560 earlier in the modules. 109 00:10:26,740 --> 00:10:31,300 We'll tell you what other methods can be employed to guarantee message not repudiation and how this 110 00:10:31,300 --> 00:10:32,530 is achieved. 111 00:10:32,800 --> 00:10:38,190 Or in other words how does a digital signature work. 112 00:10:38,270 --> 00:10:44,100 At the end we'll also discuss a way in which three basic algorithm types symmetric ciphers public key 113 00:10:44,100 --> 00:10:50,010 ciphers and hash functions are used together in modern cryptographic systems. 114 00:10:50,030 --> 00:10:51,080 See you in the training.