@ Key terms Data integrity - the ‘accuracy, completeness ‘and consistency of data, Validation - method Used to ensure entered data is reasonable and meets certain input criteria. Verification - method Used to ensure data is correct by using double entry or visual checks. Check digit - additionat digit appended to a number to check if entered data is error free. Moduto-11 - method used to calculate a check digit based on modulus division by 11 Checksum - verification method used to check if data transferred has been altered or corrupted, calculated from the block of data to be sent Parity check - method Used to check if data has been transferred correctly that uses even or odd parity. Parity bit - an extra bit found at the end of a byte that is set to “ifthe parity of the byte needs to change toagree with sender/ receiver parity protocol, Odd parity - binary number with an odd number of 1-bits Even parity ~ binary number with an even number of 1-bits. Parity block - horizontal ‘and vertical parity check ‘on a block of data being transferred. 6.2 Data integrity WHAT YOU SHOULD ALREADY KNOW Try these three questions before you read the second part of this chapter. 1. Look at the following validation screen from a spreadsheet. Why is it important to have validation in applications such as spreadsheets? Foret Celis Number Alignment Font Border Fil_—_—Protection| categonr Sample De CT . 4 Maren 2012 wosa012 vara apse Man 2012.03.14 oaale locations English (United Kingdom) Date onmats display date and time seal numbers as date values, Date formats that begin with ‘an asterisk () respond to changes in regional date and tine settings that re spectied forthe ‘operating system, Fonmats witout an asterisk ae no affected by operating system settings, 2 Why is proofreading not the same as verification? 3. Discuss one way online form designers can ensure that only certain data can be input by a user. Use the date: 12 March 2019 as the example. Data stored on a computer should always be accurate, consistent and up to date. Two of the methods used to ensure data integrity are validation and verification. The accuracy (integrity) of data can be compromised » during the data entry and data transmission stages >» by malicious attacks on the data, for example caused by malware and hacking » by accidental data loss caused through hardware issues. These risks - together with ways of mitigating them - are discussed in the rest of this chapter. 6.2.1 Validation Validation is a method of checking if entered data is reasonable (and within a given criteria), but it cannot check if data is correct or accurate. For example, if somebody accidentally enters their age as 62 instead of 26, it is reasonable ‘Asbo 819 2°96 SECURITY, PRIVACY AND DATA INTEGRITY Key terms Parity byte ~ additional byte sent with transmitted data to enable vertical parity checking fas well ‘as horizontal parity checking] to be carried out ‘Automatic repeat request (ARQ) - a type of verification check, ‘Acknowledgement - message sent toa receiver to indicate that data has been received without error. jut ~ time allowed to elapse before an ‘acknowledgement is received. but not accurate or correct. Validation is carried out by computer software; the most common types are shown in Table 6.1. Validation | Description Example of data Example of data test failing vatidation test | passing validation test type checks whether typing sk.34 in a field | typing 34.50 in a field non-numeric data which should contain | which should contain has been input into a | the price of an item | the price of an item rnumeric-only field range checks whether data | typing in somebody's | typing in somebody's entered is between | age as -120 age as 48 a lower and an upper timit format | checks whether data__| typing in the date as _| typing in the date as has been entered in the | 12-12-20 where the | 12/12/2020 where the agreed format format is dd/mm/yyyy_| format is dd/mm/yyvy length [checks whether data | typing ina telephone | typing ina telephone has the required number as 012 345 678 | number as. number of characters or | when it should contain | 012 345 678 90 when it numbers 11 digits should contain 11 digits presence | checks to make sure a | please enter passport _| please enter passport field is not left empty | number number: AB 1234567 CD when it should contain data existence | checks if data in afile | data look up for car __| data look up for a file ora file name actually | registration plate A123 | called books_in_stock exists BCD which does not | which exists in a exist database limit check | Checks only one of | typing in age as 25 | typing in somebody's the limits (such as__| where the data entered | age as 72 where the ‘the upper limit OR the | should not be negative | upper limit is 140 lower timit) consistency | checks whether data | typing in Mr in the title | typing in Ms in the check ‘in two or more fields _| field and then choosing | title field and then match up correctly | female in the sex field | choosing female in the sex field uniqueness | checks that each choosing the user name | choosing the website check entered value is unique | MAXIMUS222 in a social | name Aristooo.com networking site but | which is not already the username already | used exists A Table 6.1 Common vali 6.2.2 Verification Verification is a way of preventing errors when data is entered manually (using a keyboard, for example) or when data is transferred from one computer to another. Verification during data entry When data is manually entered into a computer it needs to undergo verification to ensure there are no errors. There are three ways of doing this: double entry, visual check and check digits.A Figure 6.6 Barcode Double entry Data is entered twice, using two different people, and then compared (either after data entry or during the data entry process). Visual check Entered data is compared with the original document (in other words, what is con the screen is compared to the data on the original paper documents). Check digits The check digit is an additional digit added to a number (usually in the right- most position). They are often used in barcodes, ISBNs (found on the cover of a book) and VINs (vehicle identification number). The check digit can be used to ensure the barcode, for example, has been correctly inputted. The check digit can catch errors including » an incorrect digit being entered (such as 8190 instead of 8180) »> a transposition error where two numbers have been swapped (such as 8108 instead of 8180) » digits being omitted or added (such as 818 or 81180 instead of 8180) phonetic errors such as 13 (thirteen) instead of 30 (thirty). Figure 6.6 shows a barcode with an ISBN-13 code with check digit. ‘An example of a check digit calculation is modulo-11. The following algorithm is used to generate the check digit for a number with seven digits: 1 Each digit in the number is given a weighting of 7, 6, 5, 4, 3, 2 or 1, starting from the left. 2 The digit is multiplied by its weighting and then each value is added to make a total. 3. The total is divided by 11 and the remainder subtracted from 11. 4 The check digit is the value generated; note if the check digit is 10 then X is used. For example: Seven digit number: 4156710 Weighting values: 7654321 Sum: (74) + (6x 1) 4 (5 5) #6 x6) +87) +(2«1)+(1 x0) = 2B +64 2542464214240 Total: = 106 Divide total by 11: 9 remainder 7 subtract remainder from 11: 11-7 = 4 (check digit) final number: 41567104 When Uhis number is entered, Une check digit is recalculated and, if Ue samme value is not generated, an error has occurred. For example, if 41576104 was entered, the check digit generated would be 3, indicating an error. ‘Asbo 819 2°96 SECURITY, PRIVACY AND DATA INTEGRITY EXTENSION ACTIVITY 6D Find out how the ISBN-13 method works and confirm that the number 978 034 098 382 has a check digit of 9. Find the check digits for the following numbers using both modulo-11 and ISBN-13. a) 213 111 000 428 b) 909 812 123 544 3 Find a common use for the modulo-11 method of generating check digits. Verification during data transfer When data is transferred electronically from one device to another, there is always the possibility of data corruption or even data loss. A number of ways exist to minimise this risk. Checksums ‘A checksum is a method to check if data has been changed or corrupted following data transmission. Data is sent in blocks and an additional value, the checksum, is sent at the end of the block of data. To explain how this works, we will assume the checksum of a block of data is ‘byte in length. This gives a maximum value of 28 - 1 (= 255). The value 0000 0000 is ignored in this calculation. The following explains how a checksum is generated. If the sum of all the bytes in the transmitted block of data is < 255, then the checksum is this value. However, if the sum of all the bytes in the data block > 255, then the checksum is found using the following simple algorithm. In the example we will assume the value of X is 1185. © (X= 1185); 185/256 = 4.609 divide the sum, X, of the bytes by 256 4 ® Rounding down to nearest whole round the answer down to the number gives Y = 4 nearest whole number, Y 4 a eng ho za" 4 The difference (X - Z) gives the ® checksum: (1185 - 1024) = 161 calculate the difference (X - 2) 4 ® This gives the checksum: 161 the value is the checksum A Figure 6.7 When a block of data is about to be transmitted, the checksum for the bytes is first calculated. This value is transmitted with the block of data. At the receiving end, the checksum is re-calculated from the block of data received. This calculated value is compared to the checksum transmitted. If they are the same, then the data was transmitted without any errors; if they are different, then a request is sent for the data to be re-transmitted.Parity checks A parity check is another method to check whether data has been changed or corrupted following transmission from one device or medium to another. A byte of data, for example, is allocated a parity bit. This is allocated before transmission. Systems that use even parity have an even number of 1-bits; systems that use odd parity have an odd number of 1-bits. Consider the following byte: 1 1 oO 1 1 0 oO parity bit A Figure 6.8 If this byte is using even parity, then the parity bit needs to be 0 since there is already an even number of 1-bits (in this case, four). If odd parity is being used, then the parity bit needs to be 1 to make the number of 1-bits odd. Therefore, the byte just before transmission would be: either (even parity): ofififovtififofo party bie or (odd parity): atijtijztof[ififofo parity bit re 6.9 Before data is transferred, an agreement is made between sender and receiver regarding which of the two types of parity are used. This is an example of a protocol. EXTENSION ACTIVITY 6E ind the parity bits for each of the following bytes: 1101101 evenparity used 0001111 — evenparity used 0111000 — evenparity used 0100 oddparity used 1 11 011011 — oddparityused If a byte has been transmitted from ‘A’ to ‘B, and even parity is used, an error would be flagged if the byte now had an odd number of 1-bits at the receiver's end. For example: Sender's byte: olifofijlijfilfojo party bit Receiver's byte: oliftofolififofo party bit A Figure 6.10 ‘Asbo 819 2°96 SECURITY, PRIVACY AND DATA INTEGRITY In this case, the receiver's byte has three 1-bits, which means it now has odd parity, while the byte from the sender had even parity (four 1-bits). This means an ettor has occurred during the transmission of the data. The error is detected by the computer re-calculating the parity of the byte sent. If even parity has been agreed between sender and receiver, then a change of parity in the received byte indicates that a transmission error has occurred. EXTENSION ACTIVITY 6F 11 Which of the following bytes have an error following data transmission? a}11101101 even parity used bho1oo01111 even parity used c}00111000 even parity used d)11110100 odd parity used el11011011 odd parity used 2 In each case where an error occurs, can you work out which bit is incorrect? Naturally, any of the bits in the above example could have been changed leading to a transmission error. Therefore, even though an error has been flagged, it is impossible to know exactly which bit is in error. One of the ways around this problem is to use parity blocks. In this method, 2 block of data is sent and the number of 1-bits are totalled horizontally and vertically (in other words, a parity check is done in both horizontal and vertical directions). As the following example shows, this method not only identifies that an error has occurred but also indicates where the error is. In this example, nine bytes of data have been transmitted. Agreement has been made that even parity will be used. Another byte, known as the parity byte, has also been sent. This byte consists entirely of the parity bits produced by the vertical parity check. The parity byte also indicates the end of the block of data. Table 6.2 shows how the data arrived at the receiving end: parity bit] bit2 | bits | bits | bits | bite | bit7 | bits bytet| 1 1 1 1 ° 1 1 0 byte2| 1 ° o 1 0 1 0 1 byte3s| 0 1 1 1 1 1 1 0 bytes] a 0 ° 0 0 0 1 0 bytes| 0 1 1 0 1 0 0 1 byteo| 1 0 0 0 1 ° 0 0 byte7| 1 0 1 0 1 1 a 1 bytes| 0 0 0 1 1 0 1 0 byte9| 0 0 0 1 0 0 1 0 parity | 1 1 0 1 0 0 0 1 byte ‘A Table 6.2A careful study of the table shows that » byte 8 (row 8) has incorrect parity (there are three 1-bits) » bit 5 (column 5) also has incorrect parity (there are five 1-bits). First, the table shows that an error has occurred following data transmission. Second, at the intersection of row 8 and column 5, the position of the incorrect: bit value (which caused the error) can be found, This means that byte 8 should have been: ofofofijfofojfilfo which would also correct column 5 giving an even vertical parity (now has four vbits). This byte could, therefore, be corrected automatically, as shown above, or an error message could be relayed back to the sender asking them to re-transmit the block of data. One Final point; if two of the bits change value following data transmission, it may be impossible to locate the error using the above method. For example, using the above example agai ofif[o[i[ififolfo This byte could reach the destination as: ofivtififtififof[i or: ofifofifofofofo or: ofif[o[ifofififo All three are clearly incorrect, but they have retained even parity so will not trigger an error message at the receiving end. Clearly, other methods to complement parity when it comes to error checking of transmitted data are required (such as checksum). Automatic repeat request (ARQ) ‘Automatic repeat request (ARQ) is another method to check data following data transmission. This method can be summarised as follows: >» ARQ uses acknowledgement (a message sent to the receiver indicating that data has been received correctly) and timeout (the time interval allowed to elapse before an acknowledgement is received). » When the receiving device detects an error following data transmission, it asks for the data packet to be re-sent. » If no error is detected, a positive acknowledgement is sent to the sender. » The sending device will re-send the data package if - it receives a request to re-send the data, or = a timeout has occurred. » The whole process is continuous until the data packet received is correct or until the ARQ time limit (timeout) is reached. >» ARQ is often used by mobile phone networks to guarantee data integrity. ‘Asbo 819 2°9