Prediction method for single event effect induced failure rate of FPGA system in radiation

Abstract

The present invention provides a method for predicting the failure rate of single-particle effects induced by radiation in an FPGA system. The specific steps are: characterizing the radiation environment of the FPGA system; determining the sensitivity of the FPGA system to single-particle effects in the radiation environment; establishing a fluence domain reliability model for the FPGA system in the radiation environment; and calculating the single-particle effect cross section σ. SEE The characteristics of the FPGA are converted into the corresponding mean free-failure fluence (MFTF). Based on the MFTF and the total number of particles per unit area in the radiation field, Φ, a reliability model for the FPGA system is established for each energy spectrum or LET spectrum within a given radiation fluence. This invention combines physical process analysis with data-driven methods to establish a fluence-domain reliability model, thereby improving the reliability of related equipment in special physical environments.

Description

Technical Field

[0001] The present invention relates to the field of radiation environment reliability, and in particular to a method for predicting the failure rate of an FPGA system induced by single particle effects in radiation. Background Art

[0002] Field-programmable gate array (FPGA)-based systems have a wide range of applications in many fields, but the use of FPGA development systems in radiation environments has the following disadvantages: Single-event upset (SEU) problem: High-energy particles in a radiation environment can strike circuit nodes in the FPGA, potentially depositing enough charge to change the state of the node, causing the FPGA to malfunction and seriously impacting system reliability. Firmware errors: In a radiation environment, single-bit SEUs in FPGA configuration elements can cause firmware errors. Logical function errors: Radiation can affect the logic circuits within the FPGA, changing the switching characteristics of the logic gates and leading to errors in the results of logical operations. Increased current leakage: Radiation can change the characteristics of semiconductor devices such as transistors in the FPGA, resulting in increased leakage current. This increase in leakage current not only increases the FPGA's power consumption, but can also cause circuit malfunctions and even interfere with other electronic components. This phenomenon is known as single-event latch-up (SEL).

[0003] Therefore, when working in a given radiation environment, the probability of single event effects (SEE) occurring in FPGAs is usually estimated using ground simulation test methods, software simulation methods, empirical formula methods, and fault injection methods.

[0004] The traditional single-event effect (SEE) analysis method primarily multiplies the total number of bits in the design by the error rate to obtain the system error rate. This method cannot effectively characterize the error rate or mean time to failure (MTTF). When applied to complex systems exposed to radiation, it cannot provide the SEE-induced failure rate of FPGAs. The resulting SEE-induced failure rate is highly skewed, making it impossible to obtain reliability data such as reliability and failure rate. The main reasons are: FPGA-based systems operate differently from memories. Furthermore, using the same multiplication factor in the trigger flip-flop (DFF) will result in extreme overestimation. Fault injection or simulation does not determine the frequency contribution to SEE sensitivity. Furthermore, various components in FPGAs are susceptible to SEUs. However, the SEU sensitivity characteristics of these components are accurately described at the per-bit level. The error rate estimate calculated at the transistor level is too fine-grained to be properly extrapolated to complex systems. Therefore, the present invention proposes a method for predicting the SEE-induced failure rate in complex FPGA systems. By establishing a fluence-domain reliability model, it addresses the reliability assessment problem under the influence of radiation processes. Summary of the Invention

[0005] In view of the problems existing in the prior art, the present invention provides a method for predicting the failure rate of single event effect induced by radiation in FPGA system, by converting the single event effect cross section σ SEE The characteristics are converted into the corresponding mean free-fault fluence MFTF and the total number of particles passing through per unit area in the radiation field Φ, and the FPGA system reliability model of each energy spectrum segment or LET spectrum segment within a given radiation fluence is established. The reliability of the FPGA system in the radiation environment is calculated, and the reliability of the FPGA system in a specified radiation environment and a specified time period is obtained, thereby solving the reliability assessment problem under the influence of the radiation process, which is of great value for improving the reliability of related equipment in special physical environments.

[0006] The present invention provides a method for predicting the failure rate of an FPGA system due to single event effects in radiation, and the specific implementation steps are as follows:

[0007] S1. Characterize the radiation environment of the FPGA system:

[0008] S11. Obtain the LET spectrum and energy spectrum according to the work requirements and mission profile of the reliability assessment of the FPGA system in the radiation environment;

[0009] S12, dividing the LET spectrum and energy spectrum obtained in step S11 into multiple energy segments and LET segments according to requirements, and calculating the particle fluence of each energy segment and LET segment within a specified time;

[0010] S2. Determine the sensitivity of the FPGA system to single event effects in a radiation environment: Apply ions with different linear energy transfers, protons with different energies, or neutrons with different energies to perform SEE tests on the energy spectrum and LET spectrum obtained in step S11, and obtain the single event effect cross sections σ corresponding to ions with different LETs, protons with different energies, or neutrons with different energies. SEE , the single event effect cross section σ SEE The expression is:

[0011] σ SEE =#errors / fluence

[0012] Where #errors is the number of failed particles, and fluence refers to the total number of particles passing through a unit area, which is usually used to describe the intensity of a particle beam or radiation field.

[0013] S3. Establish a fluence domain reliability model for the FPGA system in a radiation environment:

[0014] S31, the single event effect cross section σ obtained in step S2 SEE The characteristic is converted into the corresponding mean free-fault flux MFTF, and the specific expression is:

[0015] MFTF=1 / σ SEE ;

[0016] S32. Based on the mean free-fault fluence MFTF obtained in step S31 and the total number of particles passing through a unit area in the radiation field Φ, establish an FPGA system reliability model for each energy spectrum segment or LET spectrum segment within a given radiation fluence. The specific expression of the fluence domain reliability model of the FPGA system is:

[0017] R(Φ)=e -φ / MFTF ;

[0018] S4. Calculate the reliability of the FPGA system in a radiation environment:

[0019] S41. Calculate the particle flux of each energy spectrum segment and LET spectrum segment according to the working time of the FPGA system in the radiation environment;

[0020] S42, using the particle flux obtained in step S41 and the fluence domain reliability model obtained in step S3 to calculate the reliability of each energy spectrum segment and LET spectrum segment;

[0021] S5. Compare the reliability result obtained in step S4 with the reliability requirement in step S11 to obtain the reliability of the FPGA system within a specified time period in a specified radiation environment.

[0022] Preferably, in step S11, the mission profile includes flight altitude, flight latitude, flight longitude and flight time.

[0023] Preferably, in step S11, the LET spectrum is the radiation spectrum of heavy ions in the radiation environment, and the energy spectrum is the radiation spectrum of protons and neutrons in the radiation environment.

[0024] Preferably, in step S2, the occurrence of single event effect events in the SEE test is random, and the single event effect cross section σ SEE The value is a constant value.

[0025] Preferably, in step S3, the total number of particles Φ passing through a unit area in the radiation field is the intensity of the radiation field in the radiation environment.

[0026] Preferably, in step S5, the FPGA system specifies a time period in a specified radiation environment, which means that the types of high-energy particles in the radiation environment and their energy spectrum are determined, and the system operating time in the radiation environment is determined.

[0027] Compared with the prior art, the present invention has the following advantages:

[0028] The present invention establishes an effective fluence domain reliability model through a mapping method combined with physical process analysis and data-driven means. It has broad application prospects in many fields such as semiconductors and nuclear energy, and is of great value in improving the reliability of related equipment in special physical environments.

[0029] Reliability assessment of semiconductor devices for space applications: In space, semiconductor devices are exposed to high-energy particles such as cosmic rays. By mapping classic reliability models to the fluence domain, we can proactively assess the risk of chip failure based on fluence estimates of the space radiation environment, and more accurately predict the lifespan of these devices in space radiation environments.

[0030] Reliability assessment of semiconductor devices for aviation or ground use: Fluence-domain reliability models can help assess the reliability of electronic devices operating in areas with high radiation levels, such as aircraft cruising altitudes and in some areas with strong ground radiation. Fluence-domain reliability models can predict the probability of chip failure due to accumulated radiation over a certain period of time, allowing for the optimal scheduling of chip replacement cycles and ensuring the long-term stable operation of monitoring systems.

[0031] Photovoltaic device reliability research: Photovoltaic cells are affected by photon fluence during operation. By mapping reliability models to the fluence domain, we can analyze the reliability of photovoltaic cells under varying light intensities, or photon fluences. Fluence-domain reliability models can be used to predict the performance degradation of photovoltaic cells, providing a reference for power plant maintenance and battery replacement planning, thereby improving the overall economic benefits of photovoltaic power generation systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 This is a flow chart of a method for predicting the failure rate of an FPGA system induced by single event effects in radiation according to the present invention;

[0033] Figure 2 The atmospheric neutron energy spectrum at 41° north latitude and -74° longitude for the method for predicting the single event effect induced failure rate of an FPGA system in radiation according to the present invention;

[0034] Figure 3 The present invention is a neutron flux 4h histogram for the prediction method of the failure rate induced by single particle effect in FPGA system under radiation. DETAILED DESCRIPTION

[0035] To fully describe the technical content, objectives and effects of the present invention, the following will be described in detail with reference to the accompanying drawings.

[0036] like Figure 1 As shown in Figure 1, the prediction method for the failure rate of FPGA systems induced by single event effects in radiation is implemented as follows:

[0037] S1. Characterize the radiation environment of the FPGA system.

[0038] S2. Determine the sensitivity of the FPGA system to single event effects in a radiation environment: Apply ions with different linear energy transfers (LETs), protons with different energies, or neutrons with different energies to perform SEE tests on the energy spectrum and LET spectrum obtained in step S11, and obtain the single event effect cross sections σ corresponding to ions with different LETs, protons with different energies, or neutrons with different energies. SEE , single event effect cross section σ SEE The expression is:

[0039] σ SEE =#errors / fluence

[0040] Where #errors is the number of failed particles, and fluence refers to the total number of particles passing through a unit area, which is usually used to describe the intensity of a particle beam or radiation field.

[0041] Furthermore, in a preferred embodiment of the present invention, the single event effect (SEE) cross section is the probability per unit area of ​​a semiconductor device that a single particle, such as a heavy ion, proton, or neutron, will hit the device and produce a specific effect, such as a single event upset (SEU) or single event lockout (SEL). The SEE cross section usually varies with the energy of the particle, and particles of different energies have different effects on the device. Therefore, the occurrence of a single event effect (SEE) event in a SEE test is random, and the single event effect cross section σ is the probability that a single particle, such as a heavy ion, proton, or neutron, will hit the device and produce a specific effect, such as a single event upset (SEU) or single event lockout (SEL). SEE The value is a constant value.

[0042] S3. Establish a fluence domain reliability model of the FPGA system in a radiation environment.

[0043] S4. Calculate the reliability of the FPGA system in a radiation environment.

[0044] S5. Based on the reliability results obtained in step S4, the reliability requirements in step S11 are compared to verify whether the reliability of the FPGA system in the specified radiation environment and time period meets the system's mission requirements. Engineering decisions are then made. Specifically, engineering decisions include component replacement and solution optimization in the FPGA system.

[0045] Furthermore, the specific steps for characterizing the radiation environment of the FPGA system in step S1 are as follows:

[0046] S11. According to the work requirements and mission profile of the reliability evaluation of the FPGA system in the radiation environment, the LET spectrum and energy spectrum are obtained.

[0047] S12. Divide the LET spectrum and energy spectrum obtained in step S11 into multiple energy segments and LET segments according to requirements, and calculate the particle fluence of each energy segment and LET segment within a specified time.

[0048] Specifically, the mission profile includes flight altitude, flight latitude, flight longitude, and flight time. The LET spectrum is the radiation spectrum of heavy ions in the radiation environment, and the energy spectrum is the radiation spectrum of protons and neutrons in the radiation environment.

[0049] Furthermore, the specific process of establishing the fluence domain reliability model of the FPGA system in the radiation environment in step S3 includes:

[0050] S31, the single event effect cross section σ obtained in step S2 SEE The characteristic is converted into the corresponding mean free-fault flux MFTF, and the specific expression is:

[0051] MFTF=1 / σ SEE .

[0052] S32. Based on the mean free-fault fluence MFTF obtained in step S31 and the total number of particles passing through a unit area in the radiation field Φ, establish an FPGA system reliability model for each energy spectrum segment or LET spectrum segment within a given radiation fluence. The specific expression of the fluence domain reliability model of the FPGA system is:

[0053] R(Φ)=e -φ / MFTF .

[0054] Specifically, the total number of particles passing through a unit area in the radiation field, Φ, is the intensity of the radiation field in the radiation environment. Reliability in a radiation environment refers to the probability that a product can successfully perform its required function under specific conditions and within a specified time. Φ is used to describe the intensity of the particle beam or radiation field.

[0055] Specifically, in step S3, the fluence-domain reliability model for the FPGA system in a radiation environment is primarily based on the traditional time-domain reliability model. By comparing the similarities between reliability parameters and single-event effect parameters, the parameters of the traditional time-domain reliability model are replaced one by one with the single-event effect parameters to complete the construction of the fluence-domain reliability model. Given that single-event effects are random events, and the reliability index distribution model is a reliability model that can describe random events, the fluence-domain reliability model is constructed by replacing the traditional time-domain parameters in the reliability index distribution model with the single-event effect parameters, thereby determining system reliability within a given radiation fluence.

[0056] Furthermore, the specific process of calculating the reliability of the FPGA system in the radiation environment in step S4 includes:

[0057] S41. Calculate the particle flux of each energy spectrum segment and LET spectrum segment based on the working time of the FPGA system in the radiation environment.

[0058] S42. Using the particle flux obtained in step S41 and the fluence domain reliability model obtained in step S3, calculate the reliability of each energy spectrum segment and LET spectrum segment.

[0059] The following further describes a method for predicting the failure rate of an FPGA system caused by single event effects in radiation, in conjunction with an embodiment of the present invention:

[0060] In this specific embodiment, the method of the present invention is used to predict the single event effect-induced failure rate of an FPGA system in radiation exposure to evaluate the reliability of an avionics system based on an embedded microprocessor FPGA under a given service radiation exposure and a given service time. The specific implementation steps are as follows:

[0061] S1. Characterize the radiation environment of avionics systems based on FPGAs embedded with microprocessors:

[0062] S11. According to the reliability requirements of the avionics system based on the FPGA embedded microprocessor: at a cruising altitude of 12,000 meters at longitude -74° and latitude 41° north, the reliability within 4 flight hours reaches 3 9s, that is, 99.9%. The LET spectrum and energy spectrum are obtained, as shown in Figure 2 shown.

[0063] S12. Based on step S11, a histogram of neutron flux and energy range (E) for 4 hours is created. Considering that the main contribution of the atmospheric neutron energy spectrum to the SEE of the device in the avionics system based on the FPGA embedded microprocessor is divided into three energy ranges: thermal neutron ≤ 0.025eV, thermal neutron 1-10MeV, and thermal neutron > 10MeV, the 4-hour neutron flux of these three energy ranges is calculated respectively, as follows: Figure 3 shown.

[0064] S2. Determine the sensitivity of the FPGA system to single event effects in a radiation environment: Apply ions with different linear energy transfers (LETs), protons with different energies, or neutrons with different energies to perform SEE tests on the energy spectrum and LET spectrum obtained in step S11, and obtain the single event effect cross sections σ corresponding to ions with different LETs, protons with different energies, or neutrons with different energies. SEE , single event effect cross section σ SEE The expression is:

[0065] σ SEE =#errors / fluence

[0066] Where #errors is the number of failed particles, and fluence refers to the total number of particles passing through a unit area, which is usually used to describe the intensity of a particle beam or radiation field.

[0067] S3. Establish a fluence domain reliability model for the FPGA system in a radiation environment:

[0068] S31, the single event effect cross section σ obtained in step S2 SEE The characteristics of are converted into the corresponding mean free-fault flux MFTF, as shown in Table 1.

[0069] Table 1 MFTF corresponding to each energy

[0070] Serial number energy <![CDATA[Cross-section (cm -2 )]]> MFTF 1 ≤0.025eV 4.40E-11 2.27E+10 2 (1~10)MeV 3.10E-08 3.23E+07 4 >10MeV 8.20E-08 1.22E+07

[0071] S32. Based on the mean free-fault fluence MFTF obtained in step S31 and the total number of particles passing through a unit area in the radiation field Φ, establish an FPGA system reliability model for each energy spectrum segment or LET spectrum segment within a given radiation fluence. The specific expression of the fluence domain reliability model of the FPGA system is:

[0072] R(Φ)=e -φ / MFTF .

[0073] S4. Calculate the reliability of the FPGA system in a radiation environment:

[0074] S41. Calculate the particle flux of each energy spectrum segment and LET spectrum segment based on the working time of the FPGA system in the radiation environment.

[0075] S42. The particle flux obtained in step S41 is used with the fluence domain reliability model obtained in step S3 to calculate the reliability of each energy spectrum segment and LET spectrum segment, as shown in Table 2.

[0076] Table 2 Reliability corresponding to each energy range

[0077]

[0078]

[0079] S5. Comparing the reliability result obtained in step S4 with the reliability requirement in step S11, it is determined that the device in the avionics system based on the FPGA embedded with the microprocessor cannot meet the reliability requirement.

[0080] If the traditional single event effects analysis method is used to evaluate the reliability of an avionics system based on an FPGA embedded with a microprocessor under a given service radiation and a given service time, the reliability indicators of the FPGA avionics system are mainly estimated based on the single event effects (SEE) sensitive characteristic data of the FPGA system.

[0081] The specific implementation steps of the traditional single event effect analysis method are:

[0082] On the basis of step S1, ions with different linear energy transfer, protons with different energies, or neutrons with different energies are used to carry out single event effects, and the single event effect cross sections σ corresponding to the ions with different linear energy transfer, protons with different energies, or neutrons with different energies are obtained. SEE .

[0083] In this specific embodiment, the failure rate of the avionics system based on the FPGA embedded microprocessor and the σ of each functional module are SEE The expression is as follows:

[0084] P(fs) system ∝P Configuration +P(fs) functionLogic +P SEFI

[0085] Where fs is the operating frequency of the avionics system based on the embedded microprocessor FPGA, P(fs) system is the failure probability of the avionics system based on embedded microprocessor FPGA, P Configuration P(fs) is the SEU rate of the configuration unit in the avionics system based on embedded microprocessor FPGA. functionLogic is the SEU rate of the functional logic blocks in the avionics system based on embedded microprocessor FPGA, the functional logic blocks include sequential logic blocks and combinational logic blocks; SEFI The probability of SEFI occurring due to single event function abort is contributed by global routing and hidden logic blocks in FPGA.

[0086] σ SEE It is calculated based on the analysis of SEE test data. The device of the avionics system based on embedded microprocessor FPGA is complex and contains multiple different functional modules. Each functional module has its own σ SEE Contributes to σ SEU The dominant contribution is usually the configuration unit or functional logic unit of the avionics system based on embedded microprocessor FPGA. In the traditional single event effect analysis method, due to the untestability of the global routing SEE in the avionics system based on embedded microprocessor FPGA, its impact on σ SEE The contribution is neglected. Therefore, in determining the dominant σ SEU Then, combined with the radiation environment spectrum during service, the σ per bit is SEU Converted to λ per bit bit , change λ bitMultiplying by the total number of bits, the total number of bits in the configuration unit and functional logic block is P(fs) system .

[0087] According to the experimentally obtained σ SEE / bit, use the error rate calculation tool to obtain the error rate per bit. In this analysis method, λ bit Multiply by the number of memory bits used in the target design to obtain the error rate λ for an avionics system based on an FPGA embedded in a microprocessor system .

[0088] Through the above analysis, in the radiation environment, the λ of the avionics system based on the embedded microprocessor FPGA is system This is lower than the failure rate derived from the traditional single-event effect analysis method. Therefore, the traditional single-event effect analysis method, which multiplies the total number of bits by the error rate λ, is not suitable for predicting the failure rate of FPGA systems induced by single-event effects in radiation.

[0089] The embodiments described above are merely descriptions of preferred implementations of the present invention and are not intended to limit the scope of the present invention. Without departing from the spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should fall within the scope of protection determined by the claims of the present invention.

Claims

1. A method for predicting the failure rate of an FPGA system due to single event effects in radiation, characterized in that: The specific implementation steps are: S1. Characterize the radiation environment of the FPGA system: S11. Obtain the LET spectrum and energy spectrum according to the work requirements and mission profile of the reliability assessment of the FPGA system in the radiation environment; S12, dividing the LET spectrum and energy spectrum obtained in step S11 into multiple energy segments and LET segments according to requirements, and calculating the particle fluence of each energy segment and LET segment within a specified time; S2. Determine the sensitivity of the FPGA system to single event effects in a radiation environment: Apply ions with different linear energy transfers, protons with different energies, or neutrons with different energies to perform SEE tests on the energy spectrum and LET spectrum obtained in step S11, and obtain the single event effect cross sections σ corresponding to ions with different LETs, protons with different energies, or neutrons with different energies. SEE , the single event effect cross section σ SEE The expression is: σ SEE =#errors / fluence Where #errors is the number of failed particles, and fluence refers to the total number of particles passing through a unit area, which is usually used to describe the intensity of a particle beam or radiation field. S3. Establish a fluence domain reliability model for the FPGA system in a radiation environment: S31, the single event effect cross section σ obtained in step S2 SEE The characteristic is converted into the corresponding mean free-fault flux MFTF, and the specific expression is: MFTF=1 / s SEE ; S32. Based on the mean free-fault fluence MFTF obtained in step S31 and the total number of particles passing through a unit area in the radiation field Φ, establish an FPGA system reliability model for each energy spectrum segment or LET spectrum segment within a given radiation fluence. The specific expression of the fluence domain reliability model of the FPGA system is: R(Φ)=e -Φ / MFTF ; S4. Calculate the reliability of the FPGA system in a radiation environment: S41. Calculate the particle flux of each energy spectrum segment and LET spectrum segment according to the working time of the FPGA system in the radiation environment; S42, using the particle flux obtained in step S41 and the fluence domain reliability model obtained in step S3 to calculate the reliability of each energy spectrum segment and LET spectrum segment; S5. Compare the reliability result obtained in step S4 with the reliability requirement in step S11 to obtain the reliability of the FPGA system within a specified time period in a specified radiation environment.

2. The method for predicting the failure rate of an FPGA system due to single event effects in radiation according to claim 1, wherein: In step S11 , the mission profile includes flight altitude, flight latitude, flight longitude and flight time.

3. The method for predicting the failure rate of an FPGA system due to single event effects in radiation according to claim 1 or 2, wherein: In step S11 , the LET spectrum is the radiation spectrum of heavy ions in the radiation environment, and the energy spectrum is the radiation spectrum of protons and neutrons in the radiation environment.

4. The method for predicting the failure rate of an FPGA system due to single event effects in radiation according to claim 1, wherein: In step S2, the occurrence of single event effect events in the SEE test is random, and the single event effect cross section σ SEE The value is a constant value.

5. The method for predicting the failure rate of an FPGA system due to single event effects in radiation according to claim 1, wherein: In step S3, the total number of particles Φ passing through a unit area in the radiation field is the intensity of the radiation field in the radiation environment.

6. The method for predicting the failure rate of an FPGA system due to single event effects in radiation according to claim 1, wherein: In step S5, the FPGA system specifies a time period in a specified radiation environment, which means that the types of high-energy particles in the radiation environment and their energy spectrum are determined, and the system operating time in the radiation environment is determined.

Images