KR20110008849A - An apparatus for estimating fault coverage of embedded systems and the method thereof - Google Patents

Abstract

PURPOSE: An automated test equipment of an embedded system for reducing rate of detecting defect for using the deformity pouring experiment is provided to obtain time and cost about a bit unit fault of the embedded system. CONSTITUTION: A defect generation part(112) generates deformity to consider the rate of detecting defect evaluation. A deformity screening device(113) distinguishes the deformity based on the information about the change of the execution hunting register of the software of an embedded system and the memory value corresponding to the execution hunting register. A deformity implantation part injects the error generation deformity into the embedded system in order to distinguish the influence that the error generation deformity.

Description

An apparatus for estimating fault coverage of embedded systems and the method

The present invention relates to an apparatus and method for evaluating a defect detection rate in a digital system, and more particularly, to an apparatus and method for evaluating a bit defect detection rate in an embedded system.

If a defect occurs in the embedded system that contains the software, the defect may generate an error and directly affect the system, and may not directly affect the system, for example, because it is detected by a defect detection algorithm in the embedded system. It may not.

When a defect occurs in an embedded system, the probability of detecting the defect is called a defect detection rate, and the most widely used technique for evaluating a defect detection rate is to inject defects into the embedded system one by one so that each defect can be inserted into the system. As a method of determining the impact, this is commonly called a fault injection experiment.

Conventional techniques for evaluating the defect detection rate of an embedded system have been widely used to analyze the experimental results by performing a number of random defect injection experiments in the embedded system. However, random defect injection experiments are generally performed hundreds of thousands or millions of defect injection experiments, which has inefficiencies that require a lot of time and manpower.

A major factor that causes the inefficiency of such random defect injection experiments is the injection of defects that do not affect the operation of the embedded system. As an example, their defect injection experiments reported that 96% of defects did not affect the behavior of embedded systems [B. Nicolescu, Y. Savaria, R. Velazco, Software detection mechanisms providing full coverage against single bit-flip faults, IEEE Transactions on Nuclear Science, vol. 51, no. 6, pp. 3510-3518, December 2004]. This means that if the defect injection experiment can be performed in advance by identifying defects that do not affect the operation of the embedded system, the efficiency of the defect injection experiment can be greatly improved.

Overcoming the inefficiency of these random defect injection experiments, techniques have been developed to perform the analysis before the defect injection experiment [D. T. Smith et al., Malicious fault list generation method, US Patent 5,561,762 (October 1, 1996), Meenakshi Sekhar, Carl R. Elks, Ron L. Williams, and Barry W. Johnson, Pre-fault injection analysis for efficient fault injection in digital i & c systems, Proceedings of the sixth american nuclear society international topical meeting on nuclear plant instrumentation, control, and human-machine interface technologies (NPIC & HMIT 2009), Knoxville, Tennessee, USA, April 5-9, 2009]. However, one of these techniques relates to a technique for generating a list of defects that cause specific error outputs to be embedded in an embedded system, which has the inefficiency of analyzing as many error outputs. The other is to perform analysis based on execution trace, which reduces the number of analyzes to be performed, but has a limitation in considering defects occurring bit by bit. There is also a method to consider the defects that occur bit by bit, but this method does not consider the execution tracking, it can be said that the relative inefficiency accordingly.

It is an object of the present invention to provide an apparatus and method for evaluating a defect detection rate of an embedded system capable of acquiring a defect detection rate for a bit unit defect of an embedded system with less time and effort.

According to an aspect of the present invention, an apparatus for evaluating a defect detection rate, which is a probability of detecting a defect when a digital data value of 0 and 1 is stored in a storage device of the system as a bit unit error in an embedded system. A defect generation unit for generating a defect from the matters to be considered for the defect detection rate evaluation, and a change in the memory value corresponding to the execution trace recording of the software of the embedded system and the execution trace recording of the software. A defect sorting unit for determining whether the defects are an error generating defect that may cause an error in the embedded system or an error non-occurring defect that will not cause an error in the embedded system, and the error generating defect is embedded in the information. Defective Note to Determine Impact on System To carry out the experiment, but includes a defective injection to inject the fault occurs an error in the embedded system, the embedded system provides fault detection rate evaluation device for evaluating the defect detection rate from the fault injection test result.

According to another aspect of the present invention, a method of evaluating a defect detection rate, which is a probability of detecting a defect when an embedded system generates a defect in which digital data values of 0 and 1 are stored in a memory device of the system due to a bit unit error. In the step of generating a defect from the considerations for the defect detection rate evaluation, the information on the change of the memory value corresponding to the execution trace recording of the software of the embedded system and the execution trace recording of the software Determining from the faults whether the faults are faulty faults that may cause errors in the embedded system or non-error faults that will not cause errors in the embedded system, and how the faulty faults affect the embedded system. The number of defect injection experiments That comprising the step of injecting a fault occurs the error in the embedded system, provides the defective injection experiments embedded system fault detection rate evaluation method for evaluating the defect detection rate from the result.

Defect detection rates for bit defects in embedded systems can be obtained with less time and effort.

As the invention allows for various changes and numerous embodiments, particular embodiments will be illustrated and described in the drawings. However, this is not intended to limit the present invention to specific embodiments, it should be understood to include all modifications, equivalents, and substitutes included in the spirit and scope of the present invention.

The terminology used herein is for the purpose of describing particular example embodiments only and is not intended to be limiting of the present invention. Singular expressions include plural expressions unless the context clearly indicates otherwise. In this application, the terms "comprise" or "having" are intended to indicate that there is a feature, number, step, operation, component, part, or combination thereof described in the specification, and one or more other features. It is to be understood that the present invention does not exclude the possibility of the presence or the addition of numbers, steps, operations, components, components, or a combination thereof.

Unless defined otherwise, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art. Terms such as those defined in the commonly used dictionaries should be construed as having meanings consistent with the meanings in the context of the related art and shall not be construed in ideal or excessively formal meanings unless expressly defined in this application. Do not.

DETAILED DESCRIPTION Hereinafter, exemplary embodiments will be described in detail with reference to the accompanying drawings, and the same or corresponding components will be given the same reference numerals regardless of the reference numerals, and redundant description thereof will be omitted.

1 is a block diagram illustrating an example of an apparatus for evaluating a defect detection rate of an embedded system according to an exemplary embodiment.

Referring to FIG. 1, the embedded system defect detection rate evaluating apparatus 100 includes an injection target defect information storing unit 111, a defect generating unit 112, a defect selecting unit 113, a software execution tracking recording storing unit 114, A memory value change information storage unit 115, a target system state information acquisition unit 116, a defect injection experiment result processing unit 117, a defect injection experiment result storage unit 118, and a defect injection unit 119. Here, 'defect' means that the digital data values of 0 and 1 are stored in the memory of the digital system as an error.

The injection target defect information storage unit 111 provides information for generating a defect to be considered for evaluating a defect detection rate. Here, 'information for generating a fault' is a 'fault type' such as 0 stuck-at-0 fault, 1 stuck-at-1 fault, and bit-flip. 'Fault location' to indicate when a fault occurs, 'Fault occurrence time' to indicate when a defect occurs, 'Fault elimination time' to indicate when the defect disappears, or 'Fault duration' to indicate how long the defect lasts. have.

Table 1 shows an example of information for generating injectable defects.

Classification Information for generating defects Fault type 0 stuck, 1 stuck, bit shifted Fault location All bits in all registers
All bits in all memory areas Fault time Just before the first instruction is executed, just before the final instruction is executed Fault elimination time Immediately after execution of the last command from the time of fault occurrence

The defect generator 112 receives the information for generating a defect from the injection target defect information storage 111 and generates a defect based on the information of Table 1, for example.

The defect selector 113 is generated based on the execution trace record of the software stored in the software execution trace record storage 114 and the change of the corresponding memory value stored in the memory value change information storage 115. It is determined whether the defects may cause an error in the target embedded system 200. Hereinafter, a defect that will cause an error in the embedded system 200 will be referred to as an 'error generating defect', and a defect that will not cause an error in the embedded system 200 will be referred to as a 'error not occurring defect'.

Hereinafter, the determination of whether the defect generated by the defect generation unit 112 in the defect selection unit 113 is an error occurrence defect or a non-error occurrence defect will be described according to the present embodiment.

Table 2 shows an example of a change in the memory value according to the command of the embedded system 200.

Serial Number Memory address command R0 R1 R2 R3 0 0 0 0 0 One 300001FC mov r12, r13 0 0 0 0 2 30000200 stmdb r13!, {r11-r12, r14, pc} 0 0 0 0 3 30000204 sub r11, r12, # 0x4 0 0 0 0 4 30000208 sub r13, r13, # 0x0C 0 0 0 0 5 3000020C mov r2, # 0x6 0 0 6 * 0 6 30000210 mov r0, # 0x3 3 * 0 6 0 7 30000214 mov r1, # 0x2 3 2* 6 0 8 30000218 add r3, r2, r0 3 2 6 9 * 9 3000021C str r3, [r11, #-0x18] 3 2 6 ** 9 10 30000220 add r3, r2, r1 3 2 6 8* 11 30000224 str r3, [r11, #-0x14] 3 ** 2** 6 8 12 30000228 add r3, r0, r1 3 2 6 5 * 13 3000022C str r3, [r11, #-0x10] 3 2 6 5

Here, * indicates the first time the memory value is written, and ** indicates the time when the memory value is last used. According to the software execution trace record, the validity period of the memory value is determined, from the first writing of the value to the last using the memory value. Therefore, the time between * and ** can be defined as 'the validity period of the memory value'. For example, in the R2 register of Table 2, the validity period of the memory value is from immediately before the fifth instruction execution to just after the tenth instruction execution. In the R3 register, since writing and reading are continuously repeated, 8 All areas from immediately before execution of the 13th instruction to execution of the 13th instruction can be referred to as the valid period of the memory value.

For example, the software execution tracking record storage unit 114 stores information including the serial number of Table 2 and instructions sequentially executed, and the memory value change information storage unit 115 includes R0, R1, R2, and the like. As shown in the value change for the R3 registers, information is stored that includes a change in memory value as the instruction is executed sequentially. Here, memory refers to not only registers but also all memory areas related to defect detection rate evaluation.

As an example in which the defect generated by the defect generation unit 112 in the defect selection unit 113 determines that an error has occurred, R2 register 0 from immediately before execution of the eighth instruction to immediately after execution of the 12th instruction in the defect generation unit 112. In the case of generating zero stuck defects occurring in the second (2 ⁰ position) bits, the defect sorting unit 113 uses the information on the execution trace and the change in the memory values shown in Table 2 to determine the duration of the defect and the R2 register. The periods from the immediately preceding execution of the eighth instruction to immediately after the execution of the tenth instruction are commonly included between the validity periods of the memory values. However, by confirming that the value of the R2 register does not change because the value of bit 0 of the R2 register remains 0 during this period, the defect is determined to be an error-free defect. Accordingly, the defect injection experiment for the defect is not performed, and the result of the determination is used to evaluate the defect detection rate.

As another example in which the defect generated by the defect generation unit 112 in the defect selection unit 113 determines an error occurs defect, the defect generation unit registers the bits in the R2 register 0 bit from immediately before the eighth instruction execution to immediately after the twelfth instruction execution. In the case where the generated single-fixed defect is generated, the defect sorting unit 113 uses the information on the execution trace and the change in the memory value shown in Table 2 to determine the difference between the duration of the defect and the validity period of the memory value of the R2 register. The period from just before the 8th instruction execution to just after the 10th instruction execution is included in common. If the value of bit 0 of the R2 register changes to 1 during this period, it is confirmed that the value of the R2 register changes during the above period. Thus, the defect is determined to be an error occurrence defect. Accordingly, a defect injection experiment for the defect is performed to experimentally determine the influence of the defect on the target embedded system 200, and then evaluate a defect detection rate.

An example of how the defect sorting unit 113 determines whether the zero fixation defect generated by the defect generating unit 112 is an error generating defect is common between the duration of the zero fixing defect and the validity period of the corresponding memory value. If the result is 0, a bit-wise logical OR operation is performed on the corresponding bit values stored in the memory value change information storage unit. On the other hand, when the result is 1, the corresponding 0 fixed defect is determined as an error occurrence defect. In the case of the above example, in the memory value change information shown in Table 2, 10 immediately before execution of the eighth instruction, which is a period that is commonly included between the duration of the zero-fix fault in the 0th bit of the R2 register and the validity period of the corresponding memory value. The value of register R2 until 6th instruction execution is 6, 6, 6, which means that the value of bit 0 is 0, 0, 0. The fault is determined as a non-error fault.

An example of a method for determining whether the single fixation defect generated by the defect generating unit 112 is an error occurrence defect in the defect sorting unit 113 is common between the duration of the first fixing defect and the validity period of the corresponding memory value. For the period included as, a bitwise AND operation is performed on the corresponding bit values stored in the memory value change information storage unit. When the result is 0, the corresponding 1st defect is determined as an error occurrence defect. If the result is 1, the 1 fixed defect is determined as an error-free occurrence defect. In the case of the above example, in the memory value change information shown in Table 2, 10 immediately before execution of the eighth instruction, which is a period which is commonly included between the duration of the first fixed fault in the 0th bit of the R2 register and the validity period of the corresponding memory value. The value of register R2 until immediately after the first instruction is 6, 6, 6, which means that the value of bit 0 is 0, 0, 0, and the bitwise logical result of them becomes 0, corresponding 1 A fixation defect is determined as an error occurrence defect.

Referring back to FIG. 1, the error occurrence defects are injected into the embedded system 200 through the defect injection unit to perform a defect injection experiment. The state information about it is acquired by the target system state information acquisition unit 116 and sent to the defect injection experiment result processing unit 117.

The error-free defects are transmitted directly to the defect injection test result processing unit 117 without performing the defect injection test. The defect injection test result processing unit 117 sends the results to the defect injection test result storage unit 118. The defect injection test result storage unit 118 stores the results and evaluates a defect detection rate of the target embedded system 200 based on the test results. In more detail, the defect injection experiment result processing unit 117 is based on the determination result of the defect sorting unit 113 for the generated defect and the target system state information acquired through the defect injection experiment is embedded in the embedded system It determines whether there is an influence on the 200 and whether it can be detected, and stores the result in the defect injection test result storage unit 118.

Therefore, the embedded system defect detection rate evaluation apparatus 100 screens the defects generated by the defect generation unit 112 and the defects of the error occurrence defects and the error occurrence defects in the defect selection unit 113 and the defects of the error occurrence defects. Injection experiments are performed in real time.

2 is a block diagram showing another example of an embedded system defect detection rate evaluation apparatus according to the present embodiment.

Referring to FIG. 2, the embedded system defect detection rate evaluation apparatus 300 includes an injection target defect information storage unit 311, a defect generation unit 312, a defect selection unit 313, a software execution tracking record storage unit 314, Memory value change information storage unit 315, target system state information acquisition unit 316, defect injection experiment result processing unit 317, defect injection experiment result storage unit 318, defect screening result storage unit 319 and defect injection The unit 320 is included. 2 is a configuration in which the defect screening result storage unit 319 is added to the configuration of FIG. 1.

The same configuration is the same function, detailed description thereof will be omitted, the defect is generated in the defect generation unit 312 by using the information of the injection target defect information storage unit 311, software execution in the defect sorting unit 313 Defects generated based on the execution trace record of the software stored in the trace record storage unit 314 and the change of the corresponding memory value stored in the memory value change information storage unit 315 are generated. The process of determining the defect as the defect and storing the result in the defect screening result storage unit 319 is repeatedly performed to preferentially perform the screening process on the defects generated in the defect generating unit 312.

Then, based on the information stored in the defect screening result storage unit 319, the error occurrence defects are injected into the embedded system via the defect injection unit 320 and a defect injection experiment is performed, and the status information on the target system state It is acquired by the information acquisition unit 316 and transferred to the defect injection experiment result processing unit 317. Based on the information stored in the defect screening result storage unit 319, the defects determined as non-error-prone defects are immediately transmitted to the defect injection test result processing unit 317 without performing a defect injection test. The defect injection test result processing unit 317 stores the results in the defect injection test result storage unit 318 and evaluates the defect detection rate of the target embedded system based on the defect injection test results.

In other words, the embedded system defect detection rate evaluating apparatus 300 first performs a screening process on the defects generated by the defect generation unit 312 into an error occurrence defect and a non-error occurrence defect in the defect selection unit 313. After the defect screening process is completed, defect injection experiments are performed.

1 and 2 are examples in which the above-described modules are implemented as separate modules, but some of the above-described modules may be implemented by being separated into several modules or integrated into one module. For example, the software execution tracking record storage unit 314 (114, 314) and the memory value change information storage unit 115, 315 may be implemented in a separate form, as shown, a single integrated module It may be implemented as.

3 is a flowchart illustrating a method for evaluating an embedded system defect detection rate according to the present embodiment.

Referring to FIG. 3, an injection target defect is generated from information read by reading information on a defect to be injected from the injection target defect information storage unit (S110). Then, it is determined whether the generated injection target defect is an error occurrence defect that may cause an error in the target embedded system (S120). Next, a defect injection test for injecting an error generation defect that may cause an error in the target embedded system into the target embedded system and obtaining state information in the target embedded system is performed (S130). Then, the impact of the generated injection target defect on the target embedded system is determined based on the determination result of the generated injection target defect and the result of the defect injection experiment, and the defect detection rate is calculated and stored thereon (S140). ).

Since the steps of FIG. 3 are an example of the present invention, each of the illustrated steps may be performed separately from each other, together, or may be performed out of the order shown. In addition, some steps may be omitted or additional steps may be added as required by an operator.

Table 3 shows an example of the evaluation of the effect of the defects generated by the present embodiment on the target embedded system.

Defect type
Fault location Defect
Time Defect
time System impact Memory beat 0 stuck R1 0 0 everlasting No influence 0 stuck R1 One 0 everlasting Incorrect output 0 stuck R1 2 0 everlasting Detected 0 stuck R1 3 0 everlasting Detected 0 stuck R1 4 0 everlasting Incorrect output 0 stuck R1 5 0 everlasting Infinite loop

Table 3 shows the result of evaluating the effect of each defect on the target embedded system if there is a permanent zero-fix fault in each bit of the R1 register at the fault location before software execution. Table 3 classifies the impact on the target embedded system into four categories: no effect, detected, infinite loop, and faulty output.

Table 4 shows an example of the results of classifying the impact evaluation results for the target embedded system of the defects generated as shown in Table 3.

Classification Defect Number percentage No influence 3,992 62.38% Detected 1,023 15.98% Infinite loop 952 14.88% Incorrect output 433 6.77% Sum 6,400 100.00%

Defect detection rate evaluation for the target embedded system can be made based on the classification results as shown in Table 4. For example, 15.98% of the defects classified as detected in Table 4 as the actual detected ratio among the defects to be injected may be evaluated as the defect detection rate.

If only 5% of the total memory area is selected for defects to be injected, and if 95% of the other defects occur, it can be assumed that there will be no effect on the target embedded system. It may be evaluated as in Equation 1 below.

Therefore, the defect detection rate is 80%.

If the defect detection rate is defined as the probability of detection among defects that actually affect the system, the defect detection rate may be evaluated as shown in Equation 2 below.

According to Equation 2, the defect detection rate is 42.48%.

The defect detection rate can be defined in various ways depending on the situation. The defect injection test result processing unit evaluates the defect detection rate appropriately defined by the user according to the situation. Although the defect detection rates are defined according to different definitions, the evaluation of these defects is based on the results of the defect injection test impact classification shown in Table 4.

In the example shown in Table 4, the defects classified as No Impact account for 62.38% of all generated defects. In this embodiment, the defect injection experiment is performed only on the defects that may affect the output of the target digital system in the defect selection unit using the software execution tracking information and the memory value change information. 62.38% of the effort could be reduced. As described above, according to the present exemplary embodiment, the defect detection rate for the bit unit defect of the embedded system can be obtained even with less time and effort.

The specific numerical values used in the above-described embodiments are only for describing one embodiment of the present invention, and thus the content of the present invention is not limited to these specific numerical values.

The term '~ part' used in the present embodiment refers to software or a hardware component such as a field-programmable gate array (FPGA) or an ASIC, and '~ part' performs certain roles. However, '~' is not meant to be limited to software or hardware. '~ Portion' may be configured to be in an addressable storage medium or may be configured to play one or more processors. Thus, as an example, '~' means components such as software components, object-oriented software components, class components, and task components, and processes, functions, properties, procedures, and the like. Subroutines, segments of program code, drivers, firmware, microcode, circuitry, data, databases, data structures, tables, arrays, and variables. Functions provided within components and 'parts' may be broken down into smaller numbers of components and 'parts' or further separated into additional components and 'parts'. In addition, the components and '~' may be implemented to play one or more CPUs in the device or secure multimedia card.

All of the above functions may be performed by a processor such as a microprocessor, a controller, a microcontroller, an application specific integrated circuit (ASIC), or the like according to software or program code coded to perform the function. The design, development and implementation of the code will be apparent to those skilled in the art based on the description of the present invention.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it will be understood by those skilled in the art that various changes and modifications may be made therein without departing from the spirit and scope of the invention. You will understand. Therefore, the present invention is not limited to the above-described embodiment, and the present invention will include all embodiments within the scope of the following claims.

1 is a block diagram illustrating an example of an apparatus for evaluating a defect detection rate of an embedded system according to an exemplary embodiment.

2 is a block diagram showing another example of an embedded system defect detection rate evaluation apparatus according to the present embodiment.

3 is a flowchart illustrating a method for evaluating an embedded system defect detection rate according to the present embodiment.

<Explanation of symbols for the main parts of the drawings>

111: injection target defect information storage unit 112: defect generation unit

113: defect screening unit 114: software execution tracking record storage

115: memory value change information storage unit 116: target system state information acquisition unit

117: defect injection test result processing unit 118: defect injection test result storage unit

119: defect injection unit

Claims (14)

A device for evaluating a defect detection rate, which is a probability of detecting a defect when an embedded system has a defect in which digital data values of 0 and 1 are stored in a memory device of the system due to a bit unit error. A defect generation unit generating a defect from the matter to be considered for evaluating the defect detection rate; Whether the defects are error-producing defects that can cause an error in the embedded system from the information on the change of the memory value corresponding to the execution trace recording of the software of the embedded system and the execution trace recording of the software, or A defect sorting unit for determining whether an error-free defect does not cause an error in the embedded system; And Including a defect injection unit for injecting the error-producing defects into the embedded system to perform a defect injection experiment to determine the effect of the error-producing defects on the embedded system, Embedded system defect detection rate evaluation device for evaluating the defect detection rate from the result of the defect injection experiment. The method of claim 1, Execution tracking recording of the software is embedded system defect detection rate evaluation device that is information containing sequentially performed instructions. The method of claim 1, And wherein the change in the memory value is a change in the value for the registers. The method of claim 1, Considerations for evaluating the defect detection rate include zero fixation defects in which the digital data values are fixed to zero, one fixation defects in which the digital data values are fixed to one, and bit value conversion in which the digital data values are switched between 0 and 1. An embedded system defect detection rate evaluation apparatus including a type of a defect, a defect position indicating a bit where a defect occurs, a defect occurrence time, and a defect extinction time. The method of claim 4, wherein If the generated defect is 0 stuck defect, In the defect sorting unit, determining whether the 0 fixed defect is an error occurrence defect A period that is commonly included between the duration of the zero fixation fault and the validity period of the memory value that is the time between when the memory value is first written to the storage space into which the generated defect is to be injected and when the memory value is last used And performing a bitwise OR operation on the corresponding bit values corresponding to the change in the memory value, and when the result value is 1, the embedded system defect detection rate evaluation device for determining the 0 fixed defect as an error occurrence defect. The method of claim 4, wherein If the generated defect is 1 fixed defect, In the defect sorting unit, whether the first fixed defect is an error occurrence defect A period that is commonly included between the duration of the first fixation fault and the validity period of the memory value that is the time between when the memory value is first written to the storage space into which the generated defect is to be injected and when the memory value is last used Embedded system defect detection rate evaluation device for performing a bit-wise AND operation on the corresponding bit values corresponding to the change of the memory value to determine that the first fixed defect is an error occurrence defect. . The method of claim 1, The apparatus may further include a defect screening result storage unit configured to store the generated defect as a result of determining the error generating defect and the error non-occurring defect. And a defect injector evaluating an embedded system defect detection rate based on information stored in the defect sorting result storage unit. The method of claim 1, When the effect of the faulty fault on the embedded system is classified into (a) when the fault is detected, when the infinite loop is generated (b) and when the wrong output is generated (c), the defect detection rate is

Embedded system defect detection rate evaluation device. A method for evaluating a defect detection rate, which is a probability of detecting a defect when an embedded system has a defect in which digital data values of 0 and 1 are stored in a memory device of the system due to a bit error. Generating a defect from the points to be considered for evaluating the defect detection rate; Whether the defects are error-producing defects that can cause an error in the embedded system from the information on the change of the memory value corresponding to the execution trace recording of the software of the embedded system and the execution trace recording of the software, or Determining whether an error-free defect will not cause an error in the embedded system; And Injecting the faulty defect into the embedded system to perform a defect injection experiment to determine the effect of the faulty defect on the embedded system, Embedded system defect detection rate evaluation method for evaluating the defect detection rate from the determination result of the generated injection target defect and the defect injection test results. The method of claim 9, Execution tracking recording of the software is embedded system defect detection rate evaluation method comprising information sequentially performed instructions. The method of claim 9, And wherein the change in the memory value is a change in the value for the registers. The method of claim 9, Considerations for evaluating the defect detection rate include zero fixation defects in which the digital data values are fixed to zero, one fixation defects in which the digital data values are fixed to one, and bit value conversion in which the digital data values are switched between 0 and 1. Embedded system defect detection rate evaluation method including the type of defect, the defect position indicating the bit where the defect occurs, the defect occurrence time and the defect extinction time. 13. The method of claim 12, In the determining of whether the generated defect is an error occurrence defect or an error non-occurrence defect, If the generated defect is 0 stuck defect, The determination of whether the 0 fixed defect is an error occurrence defect is a time between the duration of the 0 fixed defect and the time when the memory value is first written to the storage space into which the generated defect is to be injected and the time when the memory value is last used. If the result value is 1, a bitwise OR operation is performed on the corresponding bit values corresponding to the change of the memory value for a period commonly included between the valid periods of the memory value. Method for evaluating the defect rate of embedded system to determine the occurrence defect. 13. The method of claim 12, In the determining of whether the generated defect is an error occurrence defect or an error non-occurrence defect, If the generated defect is 1 fixed defect, The determination of whether the first fixation fault is an error occurrence defect is a time between a duration of the first fixation fault and a time point when a memory value is first written to a storage space into which the generated defect is to be injected, and a time point when the memory value is last used. Bitwise AND operations are performed on the corresponding bit values corresponding to the change of the memory value for a period commonly included between the valid periods of the memory values. Method for evaluating the embedded system defect detection rate to determine the fault that occurred.

Images