Patent Application
20230176945
This application claims the benefit of Taiwan application Serial No. 110145674, filed Dec. 7, 2021, the disclosure of which is incorporated by reference herein in its entirety.
The disclosure relates to an error detection device and an error detection method, and relates to a device for detecting and/or correcting erroneous execution of the operation program and a method thereof.
Due to the environmental factors, the memory or register of a system platform may be interfered with or even may be damaged, and the numeric content stored in the memory or register may be changed which leads to erroneous operation result of the operation program executed on the system platform. In the space outside the atmosphere, environmental factors such as particle collision or electromagnetic interference are even worse and may easily result in single-event upset (SEU). For example, single particle in the outer space may collide with a satellite equipment operating in the outer space and change the state of the memory or register of the satellite equipment and reverse the bit data stored in the memory or register, causing execution error to the operation program of the satellite equipment and generating an erroneous operation result. Even more, the operation program may crash, and the overall task may fail.
For of the operation of the satellite equipment, due to the factors of cost, environment and distance, if errors occur to the software operation program, it is difficult to re-start the satellite equipment or re-execute the system to correct the errors of the operation program. Although the numeric errors caused by single-event upset may be resolved by upgrading hardware specification of the equipment, the cost for upgrading the specification of satellite equipment is extremely high.
Therefore, skilled ones in related industries of this technical field are devoted to resolve, using software algorithm, the technical problems caused by numeric errors arising from single-event upset among the environmental factors.
According to one embodiment, an error detection and correction device for detecting and/or correcting error(s) of an operation program is provided. The error detection and correction device includes a conversion unit, a program execution unit, a checksum operation unit, a comparison unit, and an output unit. A conversion unit is for converting the operation program into a two's complement inverse operation program, and converting a plurality of variables of the operation program into a plurality of two's complement variables. A program execution unit is for executing the operation program to obtain a first operation result according to the variables, and executing the two's complement inverse operation program to obtain a second operation result according to the two's complement variables. A checksum operation unit is for calculating a checksum result corresponding to the variables according to the operation program, and calculating a checksum result corresponding to the two's complement variables according to the two's complement inverse operation program. A comparison unit is for performing at least one of the following: comparing the first operation result with the second operation result, comparing the checksum result of the variables, or comparing the checksum result of the two's complement variables. An output unit is for outputting the first operation result, the second operation result or an error message according to a comparison result of the first operation result and the second operation result, the checksum result corresponding to the variables and/or the checksum result corresponding to the two's complement variables.
According to another embodiment, an error detection and correction method for detecting and/or correcting error(s) of an operation program is provided. The error detection and correction method includes the following steps. Converting the operation program into a two's complement inverse operation program. Converting a plurality of variables of the operation program into a plurality of two's complement variables. Executing the operation program to obtain a first operation result according to the variables. Executing the two's complement inverse operation program to obtain a second operation result according to the two's complement variables. Calculating a checksum result corresponding to the variables according to the operation program. Calculating a checksum result corresponding to the two's complement variables according to the two's complement inverse operation program. Comparing the first operation result with the second operation result, comparing the checksum result of the variables, or comparing the checksum result of the two's complement variables. Outputting the first operation result, the second operation result or an error message according to a comparison result of the first operation result and the second operation result, the checksum result corresponding to the variables and/or the checksum result corresponding to the two's complement variables.
The above and other aspects of the invention will become better understood with regard to the following detailed description of the preferred but non-limiting embodiment (s). The following description is made with reference to the accompanying drawings.
In the following detailed description, for purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of the disclosed embodiments. It will be apparent, however, that one or more embodiments may be practiced without these specific details. In other instances, well-known structures and devices are schematically shown in order to simplify the drawing.
Technical terms are used in the specification with reference to the prior art used in the technology field. For any terms described or defined in the specification, the descriptions and definitions in the specification shall prevail. Each embodiment of the present disclosure has one or more technical features. Given that each embodiment is implementable, a person ordinarily skilled in the art may selectively implement or combine some or all technical features of any embodiment of the present disclosure.
The operation program 100 has one or more variables 120. The error detection and correction device 1000 may access values of the variables 120 from the storage unit 600. Furthermore, the conversion unit 200 may perform two's complement conversion to obtain the two's complement variables 120C of the two's complement inverse operation program 100C according to the variables 120. Correspondingly, values of the two's complement variables 120C may be accessed from the storage unit 600. Moreover, the operation program 100 includes one or more functions 130. The conversion unit 200 may perform two's complement conversion to obtain the inverse operation functions 130C of the two's complement inverse operation program 100C according to the functions 130.
The program execution unit 250 may execute the functions 130 to obtain a first operation result 140 of the operation program 100 according to the variables 120. Correspondingly, the program execution unit 250 may execute the inverse operation functions 130C to obtain a second operation result 140C of the two's complement inverse operation program 100C according to the two's complement variables 120C.
The comparison unit 400 may compare the first operation result 140 with the second operation result 140C. If the second operation result 140C is equal to the two's complement of the first operation result 140, it is determined that the program execution unit 250 correctly executes the functions 130 of the operation program 100, and no numeric errors occur when the values of the variables 120 are accessed from the storage unit 600. Furthermore, it is determined that the program execution unit 250 correctly executes the inverse operation functions 130C of the two's complement inverse operation program 100C, and no numeric errors occur when the values of the two's complement variables 120C are accessed from the storage unit 600.
The checksum operation unit 300 may perform checksum operation on the variables 120 to assist determining whether the operation program 100 is executed correctly. Correspondingly, the checksum operation unit 300 may perform checksum operation on the two's complement variables 120C to assist determining whether two's complement inverse operation program 100C is executed correctly.
The output unit 500 may output the first operation result 140, the second operation result 140C or the error message 150 according to the comparison result of the comparison unit 400 and the checksum operation result of the checksum operation unit 300. If the comparison results and the checksum operation results indicate that the operation program 100 is executed correctly, the output unit 500 outputs the first operation result 140. On the other hand, if the comparison results and the checksum operation results indicate that the operation program 100 and the two's complement inverse operation program 100C are both erroneously-executed, the output unit 500 outputs the error message 150.
Taking the system platform 2000 of the satellite equipment as an example. When the satellite equipment operates in the outer space, the state of the storage unit 600 of the system platform 2000 may easily be changed due to particle collision or electromagnetic interference, hence reversing the bit data stored in the storage unit 600 (e.g., reversing from bit “1” to bit “0”) and leading to numeric errors in values of the variables 120 of the operation program 100. The error detection and correction device 1000 of the disclosure performs error detection and/or error correction on the operation program 100 executed on the system platform 2000, for the system platform 2000 of the satellite equipment. The error detection and correction device 1000 may generate a first execution area 2100 and a second execution area 2200 on the system platform 2000.
The first execution area 2100 is an area where a normal software or normal processes are executed. The operation program 100 is executed in the first execution area 2100. Moreover, the error detection and correction device 1000 performs two's complement conversion to obtain a corresponding two's complement inverse operation program 100C according to the operation program 100. The two's complement inverse operation program 100C is executed in the second execution area 2200. The second execution area 2200 is the execution area of the two's complement inverse operation process.
Basic operations of the error detection and correction device 1000 are briefly disclosed in above paragraphs. Detailed operations of the error detection and correction device 1000 are disclosed below in detail with reference to steps and flows of the error detection and correction method illustrated in
Next, at step S130, the duplicated program is translated and converted by the conversion unit 200 to obtain a two's complement inverse operation program 100C. The functions 130 of the duplicated program are respectively converted into the inverse operation functions 130C by the conversion unit 200, wherein the two's complement inverse operation program 100C is composed of the inverse operation functions 130C. Next, at step S140, the operation program 100 and the two's complement inverse operation program 100C are compiled. Next, at step S150, the compiled operation program 100 and the compiled two's complement inverse operation program 100C are linked, so as to obtain an exe file that may be executed on the system platform 2000. Next, at step S160, the operation program 100 and the two's complement inverse operation program 100C are executed on the system platform 2000, either concurrently or consecutively, according to the linked exe file, and subsequent error detection and/or error correction are performed.
Detailed implementation for step S130 of
Referring to Table 1, taking the functions 130 of “arithmetic operation” as “+”, “−”, “*”, “l”, and “%”. If the operand of the functions 130 is a variable “x”, the inverse operation functions 130C are completely identical to the functions 130. That is, the inverse operation functions 130C are “+”, “−”, “*”, “/”, and “%”. On the other hand, if the functions 130 are “+”, “−”, “*”, “/”, and “%” and the operand of the functions 130 is a fixed value “m”, the inverse operation functions 130C are “−,” “+”, “*”, “/”, and “%”. If the functions 130 are “+” and “−”, the inverse operation functions 130C are two's complement inverse operations, that is, the inverse operation functions 130C are “−” and “+”; if the functions 130 are “*”, “/”, and “%”, the inverse operation functions 130C are completely identical to the functions 130. Thus, if the operand of the functions 130 is a variable “x”, the inverse operation functions 130C are completely identical to the functions 130. If the operand of the functions 130 is a fixed value “m”, the inverse operation functions 130C could be identical to the functions 130 or the inverse operation functions 130C are the two's complement inverse operations of the functions 130.
Besides, let the functions 130 of “assign operation” be “=”, “+=”, “−=”, “*=”, “/=”, and “%=”. If the operand of the functions 130 is a variable “x”, the inverse operation functions 130C are still “=”, “+=”, “−=”, “*=”, “/=”, and “%=”, and the inverse operation functions 130C are completely identical to the functions 130. If the operand of the functions 130 is a fixed value “m”, the inverse operation functions 130C of “+=” and “−=” respectively are the two's complement inverse operations of the functions 130 of “−=” and “+=”.
Also, let the functions 130 of “pointer operation” be “&” and “*”. If the operand of the functions 130 is an integer “N” whose value is fixed, the inverse operation functions 130C are completely identical to the functions 130. If the operand of the functions 130 is a variable, such as “xPtr” and “yOffset”, the inverse operation function 130C of “*(xPtr−yOffset)” is the two's complement inverse operation of the function 130 of “*(xPtr+yOffset)”.
Furthermore, let the functions 130 of “increment/decrement operation” and “operation relation” be “++”, “−−”, “>” “<”, “<=”. If the operand of the functions 130 of “increment/decrement operation” and “operation relation” is a fixed value “m”, the inverse operation functions 130 are the two's complement inverse operation of the functions 130, that is, the inverse operation functions 130 are “−−”, “++”, “<”, “<=”, “>”, “>=”. Also, in the example of the functions 130 of “relation operation”, if the functions 130 are “==” and “!=”, the inverse operation functions 130C are still “==” and “!=”, that is, the inverse operation functions 130C are identical to the functions 130.
As disclosed above, when the conversion unit 200 of the error detection and correction device 1000 performs two's complement conversion on the functions 130 and obtains the inverse operation functions 130C, most of the inverse operation functions 130C are identical to the original functions 130. When the operand of a part of the functions 130 is a fixed value, the inverse operation functions 130C may be obtained by reversing the sign of the functions 130 or adjusting the comparison relation of value size (for example, adjusting “>” to “<”).
Detailed implementations of step S160 of
As indicated in Table 2, in the operation program 100, the operand of the functions 130 is a variable 120, which may be an input variable inputted from the outside of the program or a local variable of the program. In the present embodiment, the variables 120 of the functions 130 of “=”, “<=”, “++”, “+=” are local variables “x” and “i”. Refer to the “program code” column in Table 2. In the example where the value of the variable “i” is accumulated from “1” to “10”, the operation program 100 declares that the initial values of variables “x” and “i” both are integer “0”.
When executing the functions 130 of “=”, “<=”, “++”, and “+=”, the program execution unit 250 uses variable “i” as recursion index. When recursion is performed, the value of variable “i” is progressively increased from “1” to “10”. When the function 130 of “+=” is executed within the recursion, the value of variable “i” is accumulated to variable “x”. After 10 times of recursion, the last value of variable “x” obtained by the operation program 100 is a numeric value “55”. Meanwhile, the first operation result 140 generated by the operation program 100 is a numeric value “55”.
Moreover, the conversion unit 200 performs two's complement conversion to convert variables “x” and “i” (the variables 120 of the operation program 100) respectively into a two's complement variable “(x″)” and a two's complement variable “(i″)” (the two's complement variables 120C of the two's complement inverse operation program 100C), the two's complement inverse operation program 100C declares that the two's complement variable “(x″)” and the two's complement variable are “(i″)” and “−0” respectively. Also, the conversion unit 200 converts the function 130 of increment operation, that is, “++”, into the inverse operation function 130C of decrement operation, that is, “−−”, and when the two's complement inverse operation program 100C performs recursion, the value of the two's complement variable “(i″)” progressively decreases from “−1” to “−10”. Thus, after 10 times of recursion, the last value of two's complement variable “(x″)” obtained by the two's complement inverse operation program 100C is a numeric value “−55”. Meanwhile, the second operation result 140C generated by the two's complement inverse operation program 100C is a numeric value “−55”.
Then, the comparison unit 400 compares the first operation result 140 with the second operation result 140C. The comparison result is: the numeric value “−55” of the second operation result 140C and the numeric value “55” of the first operation result 140 have identical absolute value but have opposite sings. In other words, the numeric value “−55” of the second operation result 140C is the two's complement of the numeric value “55” of the first operation result 140.
As disclosed above, the comparison result of the comparison unit 400 shows that: the first operation result 140 of the operation program 100 is equal to the two's complement of the second operation result 140C of the two's complement inverse operation program 100C. Thus, the error detection and correction device 1000 may determine that: when the program execution unit 250 executes the operation program 100 and the two's complement inverse operation program 100C, the operation program 100 and the two's complement inverse operation program 100C both are executed correctly, and during the execution of the programs, no numeric errors occur when the values of the variables 120 of “x”, “i” and the two's complement variables 1200 “(x″)” and “(i″)” are accessed from the storage unit 600.
After step S160 of
Then, at step S318, when the two's complement inverse operation program 100C is executed by the program execution unit 250, operations of the inverse operation functions 130C are performed by the program execution unit 250 to obtain a second operation result 140C according to the two's complement variables 120. Then, at step S320, checksum operation is performed on all two's complement variables 120C (such as variables “(i″)” and “(x″)”) of the two's complement inverse operation program 100C by the checksum operation unit 300 to obtain a checksum result corresponding to all two's complement variables 120 of the two's complement inverse operation program 100C.
Then, at step S322, the first operation result 140 and the second operation result 140C are compared by the comparison unit 400 to determine whether the first operation result 140 is equal to the two's complement of the second operation result 140C and accordingly determine whether the operation program 100 and the two's complement inverse operation program 100C are executed correctly. At step S322, if the first operation result 140 is equal to the two's complement of the second operation result 140C, the determination is as follows: the program execution unit 250 correctly executes the operation program 100, the program execution unit 250 also correctly executes the two's complement inverse operation program 100C, and no numeric errors occur when the values of the variables 120 of the operation program 100 and the two's complement variables 120C of the two's complement inverse operation program 100C are accessed from the storage unit 600 by the program execution unit 250 when executing the operation program 100 and the two's complement inverse operation program 100C. Thus, it may be determined that both the first operation result 140 and the second operation result 140C are correct. Then, at step S324, the first operation result 140 is outputted and returned to the system platform 2000 by the output unit 500.
On the other hand, at step S322, if the comparison result of the comparison unit 400 shows that the first operation result 140 is not equal to the two's complement of the second operation result 140C, the determination is as follows: the program execution unit 250 erroneously executes the operation program 100 and/or the program execution unit 250 erroneously executes the two's complement inverse operation program 100C (that is, at least one of the operation program 100 and the two's complement inverse operation program 100C is executed erroneously). Meanwhile, steps S330 and S340 of
Referring to
On the other hand, at step S340 to step S344, the checksum result is analyzed in the two's complement inverse operation program 100C to determine whether the two's complement inverse operation program 100C is executed correctly. At step S340, the checksum result corresponding to all two's complement variables 120C of the two's complement inverse operation program 100C is analyzed by the checksum operation unit 300. Then, at step S342, whether the checksum result corresponding to all two's complement variables 120C of the two's complement inverse operation program 100C is equal to the mutual checksum result corresponding to the variables between the “first variable” and the “last variable” of the two's complement inverse operation program 100C is determined. If the determination in step S342 shows that the above checksum results are equal, this indicates that the two's complement inverse operation program 100C is executed correctly, and at step S344, the second operation result 140C of the operation program 100 is outputted and returned to the system platform 2000. If the determination in step S342 shows that the above checksum results are not equal, this indicates that the two's complement inverse operation program 100C is executed erroneously, and at step S350, an error message 150 is outputted to the system platform 2000.
Detailed implementation of step S316 and step S320 of
Refer to Table 3. When operation is performed on any of the variables “x” and “y” of the operation program 100 and changes the value of the variable, checksum operation must be immediately performed on the variable with value change to update the checksum operation result. Let the variable “x” and the variable “y” be taken for example. After the program code of “x=1,” is performed, the value of the variable “x” changes to “1”. After the program code of “y=2;” is performed, the value of the variable “y” changes to “2”. Thus, checksum operation must be performed on the variable “x” and the variable “y” to obtain an updated checksum result: “1{circumflex over ( )}2=3”, wherein the operation symbol “{circumflex over ( )}” represents a mutually exclusive or (XOR) operation. Then, after the program code of “y=3;” is performed, the value of the variable “y” changes to “3”. Thus, checksum operation must be performed on the variable “x” and the variable “y” to obtain an updated checksum result: “1{circumflex over ( )}3=2”.
During the execution of the operation program, the “first variable” is defined as the value of the first appearing variable among the variables of the operation program 100. According to the execution of the program code as illustrated in Table 3, the “first variable” is the numeric value “1” of the variable “x”. On the other hand, the “last variable” is defined as the value of the last appearing variable among the variables of the operation program 100. According to the execution of the program code as illustrated in Table 3, the “last variable” is the numeric value “3” of the variable “y”. The checksum result obtained by performing mutual checksum operation on the variables between the first variable value “1” and the last variable value “3” is: “1{circumflex over ( )}3=2”.
Next, referring to Table 4, checksum operations are performed on all variables “x” and “i” of the operation program 100 to obtain a checksum results corresponding to all variables “x” and “i” according to the sub-functions of “checksum{circumflex over ( )}=x” and “checksum{circumflex over ( )}=i”. For example, after the variable “x” is declared, the checksum operation of “checksum{circumflex over ( )}=x” is performed on the variable “x”, both before and after the variable “i” is accumulated to the variable “x”. Similarly, after the variable “i” is declared, the checksum operation of “checksum{circumflex over ( )}=i” is performed on the variable “i”, both before and after increment operation is performed on the variable “i”.
Additionally, the “first variable” and the “last variable” are analyzed. In the operation program 100, the “first variable” is the value of the variable which firstly appears among all variables (here, is the value of variable “i”), and the “last variable” is the value of the variable which last appears among all variables (here, is the value of variable “x”). When the execution of the operation program 100 is completed, mutual checksum operation of “checksum=i{circumflex over ( )}x” corresponding to the variable between the “first variable” (the value of the firstly appearing variable “i”) and the “last variable” (the value of the last appearing variable “x”) is performed to obtain a mutual checksum result corresponding to the variable between the “first variable” and the “last variable”.
Refer to Table 5. Checksum operation is performed on all variables “x″” and “i″” to obtain a checksum result corresponding to all variable “x″” and “i″” of the two's complement inverse operation program 100C according to the sub-functions of “checksum{circumflex over ( )}=x″” and “checksum{circumflex over ( )}=i″”. Mutual checksum operation of “checksum=i″{circumflex over ( )}x″” is performed on the variables between the “first variable” and the “last variable”.
As disclosed in above embodiments, the error detection and correction device 1000 of the disclosure converts an operation program 100 into a two's complement inverse operation program 100C, executes the operation program 100 and the two's complement inverse operation program 100C, compares the operation results of the operation program 100 with the two's complement inverse operation program 100C (the first operation result 140 and the second operation result 140C) and checks whether the first operation result 140 is two's complement of the second operation result 140C to determine whether the operation program 100 and the two's complement inverse operation program 100C are executed correctly, and determine whether numeric errors occur to the values of the variables 120 and the values of the accessed two's complement variables 120C. Furthermore, the error detection and correction device 1000 may be assisted with checksum operation to enhance the accuracy of the values of the variables 120 and the two's complement variables 120C when operation program 100 and the two's complement inverse operation program 100C are executed. Thus, the disclosure is capable of detecting and/or correcting erroneous execution of the operation program 100 by using software algorithms only and resolving single-event upset with a lower cost.
It will be apparent to those skilled in the art that various modifications and variations may be made to the disclosed embodiments. It is intended that the specification and examples be considered as exemplary only, with a true scope of the disclosure being indicated by the following claims and their equals.
| Number | Date | Country | Kind |
|---|---|---|---|
| 110145674 | Dec 2021 | TW | national |
