Method for determining single event upset impact quantification index based on periodic program updates

Abstract

The application provides a single event upset influence quantification index determination method based on program periodic updating, which comprises the following steps: for the instruction Cache in the periodic updating central processing unit (CPU), quantitatively characterizing the influence of single event upset (SEU) on the electronic system as a single event upset (SEU) sensitive area, calculating the error number of the instruction Cache output when the program remains unchanged, constructing a modified physical model of the error number of the instruction Cache output after the program periodic updating step, obtaining the error number of the instruction Cache output after the program periodic updating step, and determining the quantification index of the influence of single event upset (SEU) after the program periodic updating step. The application can realize the quantitative evaluation of the single event upset (SEU) influence alleviating effect of the program periodic updating step, effectively support the quantitative evaluation of the influence of radiation effects on electronic products in the engineering implementation process, effectively support the quantitative evaluation of the reliability of the electronic system, and the quantitative evaluation result is beneficial to the development of airworthiness certification activities.

Description

Technical Field

[0001] This invention belongs to the field of system reliability technology, and in particular, it is a method for determining the quantitative indicators of the impact of single-event upsets on electronic systems based on periodic program updates. Background Technology

[0002] High-energy particle radiation environments exist in near-Earth space (0–20 km), adjacent space (20 km–200 km), and outer space (>200 km). The types and energy spectra of these high-energy particles vary with altitude. When these high-energy particles enter semiconductor devices, they can induce single-event effects (SEE), jeopardizing the reliable operation of electronic systems, increasing the frequency or rate of system failures, and affecting system reliability indicators. As the feature size of integrated circuits becomes increasingly smaller, integrated circuits are becoming more sensitive to SEE. For integrated circuits with feature sizes smaller than 90 nm, SEE has become a major source of failure, and the failure rate induced by SEE is now greater than the sum of failure rates induced by other factors. The increasingly serious hazards of SEE have attracted attention from industries such as aerospace, aviation, communications, automotive electronics, rail transportation, power, big data, and health management.

[0003] To mitigate the harmful effects of Single Event Filtering (SEE), necessary mitigation measures should be adopted based on the characteristics of electronic devices and their SEE, such as parity checking, error detection and correction, triple modular redundancy, and periodic updates. High-energy particles incident on the CPU can easily induce Single Event Flip (SEU), a failure mode within the SEE, characterized by an unexpected change in the logic state, which can be a change from "1" to "0" or vice versa. Data and instruction cache units in the CPU are more prone to SEE than other units. Data cache SEE outputs system errors that are generally data errors and do not affect program execution; mitigation measures for data cache SEE typically include parity checking and error detection and correction. Instruction cache SEE, however, can lead to program anomalies and system malfunctions, usually with very serious consequences; mitigation measures for instruction cache SEE can only be achieved through periodic program updates.

[0004] However, current assessments of the mitigation effects of periodic program updates on SEU (Self-Effect Usage) hazards lack quantitative calculations. This hinders the quantitative assessment of radiation effects on electronic products during engineering implementation and also affects the quantitative assessment of electronic system reliability. Quantitatively calculating the mitigation effects of periodic program updates on SEU hazards is not only crucial in radiation hazard assessment but also an important component of system reliability calculations. Therefore, to address this issue, it is urgent and necessary to find a method for determining the quantitative index of single-event upset (SEU) effects based on periodic program updates, in order to quantitatively assess the mitigation effects of periodic program updates on SEU hazards and facilitate airworthiness certification activities. Summary of the Invention

[0005] This invention addresses the shortcomings of existing technologies by proposing a method for determining the quantitative indicators of single-event upset (SEU) impacts based on periodically updated instruction caches in a central processing unit (CPU). This method involves periodically updating the instruction cache, quantitatively characterizing the impact of SEUs on electronic systems as the SEU-sensitive area, calculating the number of errors output by the instruction cache before the periodic update step, constructing a corrected physical model of the number of errors output by the instruction cache after the periodic update step, obtaining the number of errors output by the instruction cache after the periodic update step, and determining a quantitative indicator of the difference in mitigation effect of SEUs before and after the periodic update step. This invention enables a quantitative assessment of the effect of periodic updates on mitigating SEUs, strongly supports the quantitative assessment of the impact of radiation effects on electronic products during engineering implementation, effectively supports the quantitative assessment of electronic system reliability, and its quantitative assessment results are beneficial for airworthiness certification activities.

[0006] This invention provides a method for determining the quantitative index of single-event upset effects based on periodically updated program updates, comprising the following steps:

[0007] S1. For the periodically updated instruction cache in the central processing unit (CPU), the impact of single-event upset (SEU) on the electronic system is quantitatively characterized as the SEU-sensitive area σ:

[0008]

[0009] Where Fluence represents the high-energy particle fluence of the incident electron system; n represents the number of single-event flips (SEUs) in the electron system;

[0010] S2. Calculate the number of errors N output by the instruction cache when the program remains unchanged:

[0011] N = Flux·T·S (2)

[0012] Where Flux represents the high-energy particle fluence rate of the incident electron system; T represents the program update period; and S represents the calculated value of the single-particle flip SEU sensitive area σ when the program remains unchanged.

[0013] S3. Construct a corrected physical model of the error count output by the instruction cache after the periodic program update step, and obtain the error count N′ output by the instruction cache after the periodic program update step:

[0014] S31. Determine the effective single-event flip SEU sensitive area S′ of the Cache at time t after the program periodically updates the step;

[0015] S32. Calculate the number of errors N′ output by the instruction cache after the program periodically updates the step:

[0016]

[0017] S4. Determine the quantitative index of the impact of the periodic program update step on the single-event flip SEU: The difference in the impact of the periodic program update step on the single-event flip SEU is characterized by the ratio of the number of errors output by the instruction cache after the periodic program update step to the number of errors when the program remains unchanged. That is, the quantitative index β of the impact of the periodic program update step on the single-event flip SEU is expressed as:

[0018]

[0019] The quantitative indicator β is used to quantitatively assess the impact of the periodic update step on the single-event flip SEU.

[0020] Preferably, in step S31, the effective single-event flip SEU sensitive area S′ is analyzed in conjunction with the program execution mode. The program execution mode includes sequential execution, loop execution, jump execution, and mixed execution containing sequential statements, loop statements, and jump statements. In order to eliminate the invalid single-event flip SEU sensitive area S0, a general functional relationship of the effective single-event flip SEU sensitive area S′ is established: S′=S-S0.

[0021] Preferably, in step S31, when the program is executed sequentially, i.e., executed step by step downwards, after the program periodic update step, within one program periodic update cycle, the single-event flip SEU sensitive area σ of the instruction cache gradually decreases from S to 0. Then, the functional relationship of the effective single-event flip SEU sensitive area S′ is:

[0022]

[0023] Preferably, the invalid single-event flip SEU sensitive area is a single-event flip SEU sensitive area that does not contribute to system failure. After a program's periodic update cycle ends, the program is reloaded, and the original erroneous program is overwritten and not executed, so there is no error output.

[0024] Preferably, in step S1, the single-event flip SEU sensitive area σ is the cross-sectional area of ​​the electronic system where a single-event flip SEU occurs, which is used to characterize the sensitivity of the electronic system to the single-event flip SEU.

[0025] Preferably, the larger the value of the single-event flip SEU sensitive area σ, the larger the physical area sensitive to the single-event flip SEU, and the more sensitive the electronic system is to the single-event flip SEU.

[0026] Compared with the prior art, the technical effects of the present invention are as follows:

[0027] 1. This invention presents a method for determining the quantitative index of single-event upset (SEU) impact based on periodic program updates. This method enables a quantitative assessment of the effectiveness of periodic program updates in mitigating SEU impacts, rather than merely a qualitative assessment, thus strongly supporting the quantitative assessment of the impact of radiation effects on electronic products during engineering implementation. Furthermore, the quantitative index β is used for the quantitative assessment of the effectiveness of periodic program updates on SEU impacts and serves as a crucial input for System Functional Safety (SEE) impact analysis in airworthiness activities. It is used to determine whether the maximum permissible probability of a non-normal event (NSEE) failure mode is less than the ratio of the FMEA rate to 10.

[0028] 2. The present invention provides a method for determining the quantitative index of single-event upset effect based on periodically updated program. For systems operating in a radiation environment, this method can effectively support the quantitative assessment of electronic system reliability, and its quantitative assessment results are beneficial to the development of airworthiness certification activities. Attached Figure Description

[0029] Other features, objects, and advantages of this application will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings.

[0030] Figure 1 This is a flowchart of the method for determining the quantitative index of single-event flip impact based on periodic program updates according to the present invention;

[0031] Figure 2 This is a schematic diagram illustrating the effect of the single-particle flip SEU sensitive area on program execution according to the present invention;

[0032] Figure 3 This is a flowchart of the NSEE failure rate control index allocation based on airworthiness safety requirements of the present invention. Detailed Implementation

[0033] The present application will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and not intended to limit it. Furthermore, it should be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings.

[0034] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0035] Figure 1 The present invention illustrates a method for determining the quantitative index of single-event upset effects based on periodically updated program updates. This method includes the following steps:

[0036] S1. For the periodically updated instruction cache in the central processing unit (CPU), the impact of single-event upset (SEU) on the electronic system is quantitatively characterized as the SEU-sensitive area σ:

[0037]

[0038] Where Fluence represents the high-energy particle fluence of the incident electron system; n represents the number of single-event flips (SEUs) in the electron system.

[0039] The single-event flip (SEU) sensitive area σ is the cross-sectional area of ​​the electronic system where a single-event flip SEU occurs. It is used to characterize the sensitivity of the electronic system to the single-event flip SEU. The larger the value of σ, the larger the physical area sensitive to the single-event flip SEU, and the more sensitive the electronic system is to the single-event flip SEU.

[0040] S2. Calculate the number of errors N output by the instruction cache when the program remains unchanged:

[0041] N = Flux·T·S (2)

[0042] Where Flux represents the high-energy particle fluence rate of the incident electron system; T represents the program update period; and S represents the calculated value of the single-particle flip SEU sensitive area σ when the program remains unchanged.

[0043] In one specific embodiment, the calculated value S of the single-event flip (SEU) sensitive area σ of the instruction cache of a certain central processing unit (CPU) is 1e-5cm. 2 The program's periodic update cycle T is 8 seconds, and the loading time is extremely short and negligible. The program executes sequentially, and the high-energy particle flux of the incident electron system is 1.7 cm⁻¹. 2 ·s -1 From expression (2), the number of errors N output by the instruction cache when the program remains unchanged is:

[0044] N=Flux·T·S=1.7·8·1e-5=13.6e-5 (8).

[0045] S3. Construct a corrected physical model of the number of errors output by the instruction cache after the program periodically updates the step, and obtain the number of errors N′ output by the instruction cache after the program periodically updates the step.

[0046] S31. Determine the effective single-event flip SEU sensitive area S′ of the instruction cache at time t after the program periodically updates the step.

[0047] The effective single-event flip (SET) sensitive area S′ is analyzed in conjunction with the program execution mode, which includes sequential execution, loop execution, jump execution, and mixed execution containing sequential statements, loop statements, and jump statements. In order to eliminate the ineffective SET SET sensitive area S0, a general functional relationship of the effective SET SET sensitive area S′ is established: S′=S-S0.

[0048] Assuming the program executes sequentially, that is, step by step downwards: the program is moved from flash memory to instruction cache, executed sequentially from the first address to the last address, the program is reloaded which is the program update, and then executed sequentially from the first address to the last address again, and so on, repeating this process continuously.

[0049] After adopting the program's periodic update procedure, within one program periodic update cycle, the single-event flip (SEU) sensitive area σ of the instruction cache gradually decreases from S to 0. Therefore, the functional relationship of the effective SEU sensitive area S′ is:

[0050]

[0051] When the program executes to address X, if an error occurs in the program residing at addresses lower than X due to a single-event flip-flop (SEU), since the program is reloaded after a periodic update cycle, the original erroneous program is overwritten and will not be executed, so there is no error output. Only programs residing at addresses higher than X will output errors due to a SEU, triggering a system failure. The SEU-sensitive area for instructions in the cache below address X is S. 1 The single-event flip (SEU) sensitive area of ​​the instruction cache with address greater than X is S. 2 ,like Figure 2 As shown, S 1 They contribute nothing to system failures and should be removed; only S 2 It can induce system failure.

[0052] S32. Calculate the number of errors N′ output by the instruction cache after the program periodically updates the step:

[0053]

[0054] In a specific embodiment, the number of errors N′ output by the instruction cache after the periodic update step can be obtained from expression (4):

[0055]

[0056] The value indicates that an error occurred once in 14705.9s, or 4.08h.

[0057] S4. Determine the quantitative index of the impact of the periodic program update step on the single-event flip SEU: The difference in the impact of the periodic program update step on the single-event flip SEU is characterized by the ratio of the number of errors output by the instruction cache after the periodic program update step to the number of errors when the program remains unchanged. That is, the quantitative index β of the impact of the periodic program update step on the single-event flip SEU is expressed as:

[0058]

[0059] The quantitative indicator β is used to quantitatively assess the impact of the periodic update steps on single-event rollover (SEU). It also serves as an important input for SEE impact analysis in system functional safety analysis during airworthiness activities, and is used to determine whether the maximum permissible probability of NSEE failure mode is less than the ratio of FMEA rate to 10.

[0060] The NSEE failure rate control index allocation process based on airworthiness safety requirements is as follows: Figure 3 As shown, the β value is helpful in determining whether the maximum allowable probability of NSEE failure mode is less than the FMEA rate / 10 indicator. If this indicator is not achieved, the system needs to take further protective measures, such as minimizing the time ratio of the internal loop program within a refresh cycle.

[0061] In one specific embodiment, the quantitative index β of the impact of the periodic program update step on the single-event flip SEU is 0.5, which is obtained by formula (5). This value indicates that the effect of the periodic program update step on the single-event flip SEU is only 50% of that when the program remains unchanged, that is, the effect of the periodic program update step on the single-event flip SEU is reduced by 50%. It can be seen that the present invention achieves a quantitative evaluation of the effect of the periodic program update step on the single-event flip SEU through formula (5).

[0062] The method proposed in this invention is for, but not limited to, periodically updating the instruction cache in the central processing unit (CPU).

[0063] This invention proposes a method for determining the quantitative indicators of single-event upset (SEU) impact based on periodically updated programs. This method enables a quantitative assessment of the effect of periodically updated programs on mitigating SEU impacts, rather than merely a qualitative assessment. It strongly supports the quantitative analysis of the impact of radiation on electronic products during engineering implementation. For systems operating in radiation environments, this method can effectively support the quantitative assessment of electronic system reliability, and its quantitative assessment results are beneficial to the development of airworthiness certification activities.

[0064] Finally, it should be noted that the above embodiments are for illustration only and not for limiting the technical solutions of the present invention. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the present invention without departing from the spirit and scope of the present invention. Any modifications or partial substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for determining the quantitative index of single-event reversal impact based on periodically updated program, characterized in that, It includes the following steps: S1. For periodically updated instruction caches in the central processing unit (CPU), the impact of single-event upset (SEU) on electronic systems is quantitatively characterized as the SEU-sensitive area. : （1）； Where Fluence represents the high-energy particle fluence of the incident electron system; n represents the number of single-event flips (SEUs) in the electron system; S2. Calculate the number of errors N output by the instruction cache when the program remains unchanged: （2）； in, The high-energy particle fluence of the incident electron system is represented by T; the program update period is represented by S; and the single-event flip SEU sensitive area is represented by S when the program remains unchanged. The calculated value; S3. Construct a corrected physical model of the error count output by the instruction cache after the periodic program update step, and obtain the error count output by the instruction cache after the periodic program update step. : S31. Determine the effective single-event flip SEU sensitive area of ​​the instruction cache at time t after the program periodically updates the step. ; S32. Calculate the number of errors output by the instruction cache after the program periodically updates the steps. : （4）； S4. Determine the quantitative indicator of the impact of the periodic program update step on the Single Event Flip (SEU): The difference in the impact of the periodic program update step on the SEU is characterized by the ratio of the number of errors output by the instruction cache after the periodic program update step to the number of errors when the program remains unchanged. This is the quantitative indicator of the impact of the periodic program update step on the SEU. Represented as: （5）； Quantitative indicators Used for quantitative evaluation of the impact of periodic program update steps on single-event flip SEU; In step S31, the effective single-particle flip SEU sensitive area The analysis considers the program execution methods, which include sequential execution, loop execution, jump execution, and mixed execution containing sequential statements, loop statements, and jump statements. This is done to eliminate invalid single-event flip SEU sensitive areas. Establish an effective single-event flip SEU sensitive area The general functional relationship is: .

2. The method for determining the quantitative index of single-event reversal impact based on periodically updated program according to claim 1, characterized in that, In step S31, when the program is executed sequentially, i.e., executed step by step downwards, after the program periodic update step, within one program periodic update cycle, the single-event flip SEU sensitive area of ​​the instruction cache is... As S gradually decreases to 0, the effective single-event flip SEU sensitive area is... The functional relationship is: （3）。 3. The method for determining the quantitative index of single-event reversal impact based on periodically updated program according to claim 1, characterized in that, The invalid single-event flip SEU sensitive area is the single-event flip SEU sensitive area that does not contribute to system failure. After a program's periodic update cycle ends, the program is reloaded, and the original erroneous program is overwritten and not executed, so there is no error output.

4. The method for determining the quantitative index of single-event reversal impact based on periodically updated program according to claim 1, characterized in that, In step S1, the single-particle flip SEU sensitive area This represents the cross-sectional area of ​​the single-event upset (SEU) in an electronic system, used to characterize the sensitivity of the electronic system to the SEU.

5. The method for determining the quantitative index of single-event reversal impact based on periodically updated program according to claim 4, characterized in that, The single-particle flip SEU sensitive area The larger the value, the larger the physical area that the single-event flip SEU is sensitive to, and the more sensitive the electronic system is to the single-event flip SEU.

Images