TW200427236A - Encoding technology to detect and correct error - Google Patents

Abstract

The present invention provides an encoding technology to detect and correct errors, wherein the parity check sum that is simple for computation is mainly used and varied for application. The data string is projected from a single sequence into a two-dimensional space (byte position-bit position), so that each bit becomes one point in the two-dimensional space. When a single error occurs, only one point appears in the two-dimensional space and is corrected. If there are two errors, two points appear in the two-dimensional space and are detected; thereby the purpose of correcting a single error and detecting double errors is achieved.

Description

200427236 V. Description of the invention (1) "Technical Field" The present invention is an application of "Mm" to use the string of resources for two purposes: to calculate a simple logical check code, and to add errors to the double-error space. Single, "Prior Art" produces • Error J, which causes no corrections for 1 C a 1 line debugging t & R when stored in the Zenglikou network system 5 actually helped to set the wrong recipe 5 correct 9 The method of operation (9 times): There are many reasons, and the aim is: 2 rounds of errors: ringing, and then multiplying various kinds of failures: transmissions on the network are mostly CR = wrong technology, Redundancy C he again 'Action, and the CRC technique can only detect the error detection capability of the Er Shang ^ technique, so it must cooperate with heavy errors and does not have the ability to correct' 不 不 浪 # Hi μ Λ &), and make ^: easier増 加 网 etransm…, I. (T… 〇u… The so-called backup file, which is extremely wasteful of space, the storage in the case system is a forward broadcast error, • it is a backup error, then when the story is over There is a so-called double share preparation #: 2: Instead, it helps the law, its waste of space is worthless = the force of wrong judgment is 1 'sentence', and it can only reach a single wrong debt: Shi, ▲, For example, the above CRC technology's error detection method, time = positive; efficiency), choose between space (memory), provincial gate a (computation is often impractical due to complicated calculations, and is often criticized by people. Load (Ding raf

Page 6 200427236

So 'how to add the above-mentioned defects to the technical difficulties. Exclude, that is, the invention of this case

"Invention"

The main purpose of the present invention is to provide code checking and detectable changes. The secondary purpose of the present invention is to provide sequence projection to the second degree space, so that every point can reduce the space (memory) load. Coding for correcting and correcting errors Another object of the present invention is to provide a coding technique that can be used when a single error is detected. A coding technique that uses simple logic to detect and correct errors. A technology that uses data strings from single-bits to two-dimensional space without causing negative network problems. A type that uses the change of the logic check code to detect, correct, and correct errors is mainly a simple operation of the parity check code (Parity Check Sum), which is used to change and project a data string from a single sequence Second degree space

BY 1 I PfSlti〇n Bit P〇sitio ^), each bit becomes a point in the second degree space, when a single error occurs, 'there is only _ point in the -degree space, & yes, it can be corrected, When a double error occurs, two points appear in the second degree space, which can be detected according to the scheme, while the reachable single-error is corrected, and the double error is the purpose of detection. "Embodiment" For the convenience of your examiner to understand other features of the present invention

Page 7 19.4 200427236 V. Description of the invention (3), advantages and achieved effects can be more apparent, and the present invention is described in detail in conjunction with the 'figure' as follows: Please refer to FIG. 1, the present invention provides / A kind of "encoding technology that can detect and correct errors", which uses simple logic check codes to calculate and apply changes, and can be corrected in single error and double error detection, which mainly includes the following steps: a data input Step 1, operation step 2, conversion projection step 3, comparison step 4, correction step 5 and detection step 6; 俾 When you want to detect and correct errors, perform the following steps in order: φ The data input step 1 ·· Receive / read data string (for example, the data string is: D〇, D1, D2D3, D4, D5, D6, D7, 〇8; of which 〇〇, 01, 〇 02, 03 4, 〇5, 〇6 are user data), and perform operation step 2; this operation step 2: perform the traditional logic check code operation to obtain the D 7 'D 8 value (as shown in Figure 2), and then perform the conversion Projection step 3; where the D 7 value is determined by the traditional Edit check code plus rotation (R o ta te), is six times (feature 0137, 6 & 011) operation, the 08 value is also a traditional logical check code, seven X 0R operation; the Conversion projection step 3: Project the data string (D ata S tr # ing) from a single sequence to the second degree space (Byte P o sition Bit Position), so that each bit becomes a point in the second degree space. After obtaining the RC and XC values (as shown in Figure 3), perform the comparison step 4; the comparison step 4: the D7, D8 values and RC values, XC values

200427236 V. Description of the invention (4) Row comparison. Please refer to the double error. When the data ratio c = 〇 and X, the data will appear again. Please refer to the different comparisons. When D 〇, there is only one bit C. The value is related to XC. There is only one error between D and D because D is wrong. Therefore, when X7 is being executed, the value of XC is only sent to bits 8 to 0 of the post-execution 8 value. Therefore, the correct D operation of the X operation is shown in the figure. The RC and XC shown in Figure 4 have no error after the C value = 0, and the blank is blank, that is, Figure 4 is combined with the Error result: D 7 occurs when the unit is non-zero. When it is affected, there is a bit line correction step. When the data is correct or The various changes of the single value are detailed "(N oE rrf if viewed from the second space, the data is not normal as shown in Table T; and as shown in Figure 2, a single difference occurs). 2 i: R c value error occurs from D 0 to! 3 7: Because it is a single error, it is affected, and it can be judged that a single error occurs. 4 2 2: R 〇 value = 〇, an error occurs when D 8 , So it is not affected, but is affected by the bits. The formula of the C value of D 8 is only 0 R (D 8 value and X c value are non-zero. When the C value obtained by the single 7 operation is only one, the X 8 value can be compared by X5; an error or the following: r): R has an error 4 2 (S setting, causing XC The value of R is incorrect because of the RC value, but the XC value is the single value of D 0 D 8 because of the RC value.) After comparison, please refer to Figure 4 and cooperate with Figure 43 (DuobleErr is set to cause different comparisons). Result: When D0 to D7 occur two (one) (2r), an error occurs. At this time, '俾 when a double error occurs, depending on the bit in which the error occurs', two kinds of errors may occur.

19κ $ 9 pages 200427236 V. Description of the invention (5) Different situation: One RC value is double error 4 3 1: RC value = 0, due to the operation of rotation (Ro tate) in the formula of RC value, each member comes from different Bit, and in the formula of XC value, each member is from the same bit (Bi t Position), so when two wrong bits fall on the same RC value, they will fall on different bits of XC value, XC value There are also two bits non-zero, and the detection step 6 is performed; two RC values are double error 4 3 2: In the formula, the RC value has two bits non-zero, and the detection step 6 is performed; The error falls between D 0 and D 7, and the other error falls on D 8. When an error occurs, the RC value has a non-zero bit. According to the XC value, it can be divided into the following two cases: An XC value has double errors 4 3 3: XC value = 0, execute detection step 6; two XC values are double error 4 3 4: XC value has two bits non-zero, execute detection step 6; two errors occur at D8 435: RC value = 〇 , XC value has two bits non-zero, perform detection step 6;

The correction step 5: directly perform the corrective action; P The detection step 6: Detect the double error, that is, send a resend signal and request to resend the data string; Through the above induction, the data can be based on the change of the RC value and the XC value Perform debugging and correction; with the use of the above steps, users can

Page 10 200427236

Understand the type of error and proper handling of Rq. The double error includes the relationship between the 雔 M value and the XC value, and the probability of occurrence is an exponential sequence. When a single error bit is sealed, it is 1 / 10,000. The probability of one and three errors is one million. When the probability of double errors is over one, the probability of occurrence of more than three errors is cleaned up. 'It is obtained from the characteristics of the exponential sequence. In the calculation of 眛 —, it is one of the knives, so the memory of the present invention) is also very low, and it is not very fine and fast. To make the problem too large, to make the present invention more obvious. To make a comparative analysis as follows: Its progress and practicability are the same as the lack of habitual use 1. It wastes space and time. 2. Double faults occur at the same place. It is easy to cause misjudgment. 3. No ability to correct errors. 4. Not practical. 5. Lack of industrial competitiveness. Advantages of the invention 1. Can correct single errors and detect double errors. 2. Reduce the use of space Without negative consequences 3, the data string to the second time sequence projected by a single space, calculates faster.

Page 11 200427236 V. Description of the invention (7) 4. Wide application. 5. It is not easy to cause misjudgment. 6, very practical and progressive. 7, with industrial and commercial and industrial use value. In summary, the present invention has indeed achieved the desired effect under the previous technical structure, and it is not easy for those skilled in the art to think about it. Furthermore, the present invention has not been disclosed before the application, and its features The progress and practicability have clearly met the requirements for applying for an invention patent, and the invention application has been filed according to law

Page 12 200427236 Brief Description of the Drawings Graphic Description The first drawing is a schematic flow chart of the method of the present invention. The second figure is a schematic diagram of the operation results of the operation steps of the present invention. The second diagram is a schematic diagram of the calculation result of the conversion projection step of the present invention. The third diagram is a single error comparison diagram of the present invention. The third figure is a schematic diagram of the double error comparison of the present invention. The fourth diagram is a comparison diagram of the change of the RC value and the XC value of the present invention. Figure No. Description 1 • Data entry step 2 • Calculation step 3 • Conversion projection step 4 • Comparison step 4 1 No error 4 2 Single error 4 2 1 D 0 to D 7 Single error 4 2 2 D 8 Single error 4 3 • Double error 4 3 1 _ One RC value double error 4 3 2 Two RC value double error 4 3 3 One XC value double error 4 3 4 Double XC value double error

Page 13 200427236 Schematic description 4 3 • · D 8 Two errors occurred Corrective steps Detection steps Page 14

Claims (1)

200427236 6. Scope of patent application 1. A "encoding technique that can detect and correct errors", which uses simple logical check codes for calculation and can be changed and applied. Coding techniques that can be corrected for single errors and double error detection are mainly composed of There are the following steps: a data input step, a calculation step, a conversion projection step, a comparison step, a correction step, and a detection step; the data input step: receiving / reading a data string, and performing a calculation step; the Operation steps: Perform the traditional logic check code operation to obtain the D 7 and D 8 values, and then perform the conversion projection step; where the D 7 value is the traditional logic check code plus rotation for six times (Ro tat e & X〇R) operation, the D 8 value is also a traditional logical check code, seven X〇R operations; the conversion projection step: the data string is then projected from a single sequence into the second degree space, so that each bit Both become a point in the second degree space. After obtaining the RC value and the XC value, the comparison step is performed; the comparison step: the D 7, D 8 values are compared with the RC, XC values; the correction Steps of: performing a direct correction of the operation; the detected the steps of: detect double errors known, i.e. signal sent resend request < 1 resend data string. 2. According to a "detectable and correctable error coding technology" described in item 1 of the scope of the patent application, the comparison step can be compared to the following situations: No error: RC value = 0 and XC value = 0, If it comes in second degree Page 15 2UU42 / Zib 6. The scope of patent application is not abnormal when looking at 5. Single error. When D 0 has a pair of non-XC values, there will only be one pair. After that, D 8 values will be non-zero. 0 to D 7 So the result of the XC value calculation is exactly the double error result of D 8 value: when D 〇 is different, an R is in the rotation formula, each element falls in the same — there are two bits and two R errors will appear in one piece. A blank indicates that the data is positive: according to the error to D 7 zero 'When a single order is generated in the affected units, when only one comparison X can be compared in the single calculation, a single error can occur by mistake, because it is affected by Impact,% y An error: An error was sent, so if each bit is sent by the C value, it will be executed according to the position of the error. This will cause an error: R will be different from D 〇 Result: c value and X c value Both are only D7. Because of the RC value and a single error, the RC value and XC can determine the occurrence of a single error. Rc value = born in D 8 is affected, affects, D type, the position of the difference correction step 0, but The XC value is only one, because the RC value is determined by D Because it is a single error, the value of 8 by DO to D7 D 8 only values, n-9, resulting in two different alignment error junction to D 7, at this time, may occur in two ways: the value of? Error: rca = 〇 As a parent member, all members come from the same bit R c value, will fall on x. The value is non-zero, and the detection step is performed. The C value is double error. In the formula, the rC value formula consists of the same bit and the XC value, so when the two error bits are different, the xc value and R c value are different. There are two bits that are not 200427236. 6. The scope of patent application is zero, and the detection step is performed. When one error falls on D 0 to D 7, and the other error falls on D 8, there is one bit in the RC value. Non-zero, according to the XC value, it can be divided into the following two cases: one XC value double error: XC value = 0, performing detection steps; two XC value double errors: XC value has two non-zero bits, detection is performed Step; Two errors occur at D8: At this time, the RC value = 0, and the XC value has two bits that are non-zero, and the detection step is performed. Page 17