Method for fast evaluation of soft error rate induced by single event based on analytical function library

Abstract

This invention discloses a rapid assessment method for soft error rate caused by single-event effects based on an analytical function library. It addresses the problems of low efficiency and difficulty in supporting large-scale circuit evaluation in traditional TCAD simulation. The method constructs and calibrates device models based on process design kits to obtain accurate electrical characteristics of the calibration device model. Based on multidimensional key irradiation variables, dense TCAD sampling and parametric fitting are performed on the calibration device to construct a response function library. Based on the actual integrated circuit layout, particle transport simulation under radiation conditions is performed in Geant4 to extract the secondary particle set. The corresponding transient current is rapidly synthesized by querying the response function library and interpolating. The current is injected into the SPICE circuit netlist for simulation to determine whether single-event upset has occurred and to calculate the soft error rate. This method achieves an order-of-magnitude improvement in simulation efficiency while maintaining accuracy, making it suitable for rapid radiation resistance assessment and design hardening of complex integrated circuits.

Description

Technical Field

[0001] This invention relates to the field of integrated circuit reliability design and simulation verification technology, and in particular to a method for rapid evaluation of soft error rate caused by single-event effects based on an analytical function library. Background Technology

[0002] As semiconductor technology continues to shrink to nanometer and sub-nanometer nodes, the critical charge of memory nodes and logic cells in integrated circuits decreases dramatically, making them extremely sensitive to single-event effects induced by high-energy particles such as atmospheric neutrons and cosmic rays. Single-event upsets, as the most prevalent form of soft error, have become a key technological bottleneck restricting the development of highly reliable electronic systems.

[0003] Currently, simulation evaluation of single-event effects mainly relies on two technical approaches: one is simplified models based on experience or analysis, such as the rectangular parallelepiped model; the other is high-precision Monte Carlo methods based on full-physics numerical simulation. The former is computationally efficient, but its physical mechanisms are coarsely characterized, making it difficult to accurately simulate complex effects such as three-dimensional electric field distribution, charge sharing, and parasitic bipolar amplification in nanodevices, resulting in severely insufficient prediction accuracy. The latter, represented by the Geant4-TCAD-SPICE serial co-simulation, can achieve high-precision evaluation, but its simulation process heavily relies on computationally intensive Three-Dimensional Transient Simulation using Technology Computer-Aided Design (TCAD) three-dimensional transient simulation, leading to low efficiency and difficulty in supporting the rapid evaluation needs of large-scale circuits and massive particle events, creating a prominent contradiction between accuracy and efficiency. Geant4 is a radiation environment and nuclear reaction simulator, TCAD is a device physics simulator, and SPICE is a circuit performance simulator.

[0004] In the traditional full-physics simulation process, each radiation event requires triggering a time-consuming TCAD transient simulation to generate the device current response. This results in a linear relationship between the total simulation time and the number of radiation events, leading to extremely low computational efficiency and making it impossible to support the engineering-oriented rapid evaluation needs of large-scale circuits and massive particle events. Summary of the Invention

[0005] This invention provides a method for rapidly assessing the soft error rate caused by single-event effects based on an analytical function library. This solves the problem that the computational efficiency of TCAD three-dimensional transient simulation is extremely low in the prior art, which cannot support the engineering needs of rapid assessment of large-scale circuits and massive particle events. It achieves an order-of-magnitude improvement in simulation efficiency, thereby meeting the practical engineering needs of rapid and reliable radiation effect assessment and hardening design optimization for complex integrated circuits.

[0006] This invention provides a fast method for evaluating the soft error rate caused by single-event effects based on an analytical function library, comprising: Based on the process design kit, a sensitive device model and a corresponding SPICE circuit model are constructed. The SPICE circuit model is then used to perform electrical calibration on the sensitive device model to obtain a calibration device model with accurate electrical characteristics. Based on the multidimensional key irradiation variables that affect transient current, TCAD transient simulation under dense sampling conditions is performed on the corrected device model, and the transient current waveform obtained from the simulation is parametrically fitted based on physical mechanisms to construct a transient current response function library that correlates irradiation conditions with transient current parameters. Based on the actual layout of the integrated circuit, three-dimensional particle transport and nuclear reaction simulations under real radiation environment were performed in Geant4 to obtain the set of secondary particles of the correction device model under a single simulated event. Using the set of secondary particles as a query condition, the transient current response function library is queried and interpolated to obtain the corresponding transient current; The transient currents corresponding to each secondary particle are used as PWL current sources. Transient simulation is performed by directly injecting the circuit netlist constructed based on the SPICE circuit model into the circuit to be evaluated. Based on the simulation results, it is determined whether a single-event upset soft error has occurred, and the soft error rate of the circuit is calculated based on the statistical results.

[0007] In one possible implementation, the step of constructing a sensitive device model and a corresponding SPICE circuit model based on a process design kit, and then using the SPICE circuit model to perform electrical calibration on the sensitive device model to obtain a calibrated device model with accurate electrical characteristics, includes: Based on the process design kit and manufacturing parameters of the target integrated circuit, a sensitive device model is built in TCAD software, and a corresponding SPICE circuit model is built in SPICE software. The key static electrical characteristics of the constructed sensitive device model and the SPICE circuit model are simulated to obtain the corresponding simulation characteristic curves and electrical simulation characteristic curves. By comparing the electrical simulation characteristic curve with the simulation characteristic curve, the parameters of the sensitive device model are adjusted to obtain a calibration device model with accurate electrical characteristics.

[0008] In one possible implementation, the TCAD transient simulation of the corrected device model under dense sampling conditions is performed based on multidimensional key irradiation variables affecting the transient current, and the simulated transient current waveform is parametrically fitted based on physical mechanisms to construct a transient current response function library that correlates irradiation conditions with transient current parameters, including: A multidimensional key irradiation variable affecting transient current is defined, and a sampling space is defined based on the multidimensional key irradiation variable; wherein, the multidimensional key irradiation variable includes: particle incident position, linear energy transfer value and incident angle; The sampling density in the sampling space is determined based on prior simulation, and the transient response current waveform of the correction device model under the irradiation conditions corresponding to the sampling point is obtained by TCAD simulation at each sampling point. The transient response current waveform is parametrically fitted using a device response model based on physical mechanisms to obtain a current parameter vector with device physical meaning; Establish the mapping relationship between the multidimensional key irradiation variables and the corresponding current parameter vectors to obtain a transient current response function library that associates the multidimensional key irradiation variables with transient current parameters.

[0009] In one possible implementation, the physical mechanism-based device response model is expressed as: ; in, Indicates the carrier drift velocity; Indicates linear energy transfer value; This represents the minimum distance a charge carrier must travel from its generation to its collection. Indicates carrier lifetime; Represents the bipolar diffusion coefficient of charge carriers; It represents the projected distance of the line segment AP formed by the particle incident point A and the carrier collection point P on the particle incident trajectory; This represents the projected distance of the line segment BP formed from the exit point B to the carrier collection point P on the particle's incident trajectory. Represents the Gaussian error function; Represents the carrier collection point P Axis coordinate values; Represents the carrier collection point P Axis coordinate values; This represents the time variable calculated from the point of incidence A.

[0010] In one possible implementation, the actual layout based on the integrated circuit is used to perform three-dimensional particle transport and nuclear reaction simulations under a real radiation environment in Geant4, obtaining a set of secondary particles acting on the correction device model by a single simulated event, including: Based on the actual layout of the integrated circuit, the geometry and material composition of the devices involved in the simulation are defined, and the particle transport simulation environment is configured in Geant4 using the defined geometry and material data; Based on the particle energy spectrum of the target radiation environment, a Monte Carlo simulation is performed to track the particle incident, nuclear reaction, and secondary particle transport processes within the model of the correction device. Extract and record the set of secondary particles generated within the sensitive volume of the calibration device model during each simulation event.

[0011] In one possible implementation, the step of using the secondary particle set as a query condition to perform querying and interpolation in the transient current response function library to obtain the corresponding transient current includes: Determine the number of secondary particles in the set of secondary particles; If the number of secondary particles is equal to 1, then the information of the secondary particles is used as a query condition to perform matching and multidimensional interpolation in the transient current response function library to obtain the corresponding transient current parameter vector, and the reconstructed transient current waveform is used as the corresponding transient current. If the number of secondary particles is greater than 1, then for each secondary particle, the information of the secondary particle is used as a query condition to perform matching and multidimensional interpolation in the transient current response function library to obtain the transient current parameter vector corresponding to each particle; the corresponding transient current waveform under the individual action of each particle is reconstructed; all the reconstructed transient current waveforms are linearly superimposed, and the superimposed total waveform is used as the corresponding transient current.

[0012] In one possible implementation, determining whether a single-event upset soft error has occurred based on simulation results includes: During the SPICE transient simulation, the voltage response of key nodes in the netlist of the circuit to be evaluated is monitored; The voltage response is compared with a preset logic threshold; If the voltage response exceeds the logic threshold and causes the logic state to flip, a single-event upset soft error is determined to have occurred.

[0013] In one possible implementation, determining whether a single-event upset soft error has occurred based on simulation results includes: Obtain the static noise margin of the critical nodes in the network table of the circuit to be evaluated; wherein, the static noise margin is the maximum transient current peak or charge that the critical nodes in the network table of the circuit to be evaluated can withstand without logic flipping; The peak value or charge amount of the transient current is compared with the static noise tolerance; If the peak value or charge of the transient current exceeds the static noise tolerance, a single-event upset soft error is determined to have occurred. If the peak value or charge of the transient current does not exceed the static noise tolerance, then it is determined that no single-event upset soft error has occurred.

[0014] In one possible implementation, the formula for calculating the soft error rate of the circuit based on statistical results is as follows: ; in, Indicates the number of soft errors; This indicates the total betting volume.

[0015] One or more technical solutions provided in this invention have at least the following technical effects or advantages: This invention constructs a sensitive device model and a corresponding SPICE circuit model based on a process design kit. The SPICE circuit model is then used to electrically calibrate the sensitive device model, resulting in a calibrated device model with accurate electrical characteristics. This precise electrical calibration ensures high electrical realism for the device model used in subsequent physical simulations, laying a reliable foundation for the physical accuracy of the entire evaluation process. Based on multidimensional key irradiation variables affecting transient current, TCAD transient simulations are performed on the calibrated device model under intensive sampling conditions. The resulting transient current waveforms are then parametrically fitted based on physical mechanisms to construct a transient current response function library that correlates irradiation conditions with transient current parameters. This step, through systematic intensive sampling and parametric fitting in the early stages, solidifies the time-consuming three-dimensional TCAD transient simulation into an efficient "irradiation condition-current parameter" lookup library. This removes the most computationally intensive steps from the subsequent event loop, achieving structural optimization of the simulation process and creating conditions for orders-of-magnitude improvements in efficiency. Based on the actual layout of integrated circuits, three-dimensional particle transport and nuclear reaction simulations under realistic radiation environments are performed in Geant4 to obtain the set of secondary particles acting on the calibration device model for a single simulated event. This step, through accurate geometric modeling and Monte Carlo simulation based on the actual layout, can realistically reproduce the interaction between particles and matter in complex radiation environments, accurately acquiring all secondary particle information generated within the sensitive volume of the device for each radiation event, ensuring the physical completeness and environmental realism of the evaluation input. The set of secondary particles is used as a query condition in the transient current response function library for querying and interpolation to obtain the corresponding transient current. This step utilizes a pre-built response function library, replacing traditional real-time TCAD simulation with fast mathematical queries and multi-dimensional interpolation, achieving "on-demand synthesis" of transient current waveforms, completely solving the efficiency bottleneck of the TCAD step in traditional serial simulation, and making it possible to handle massive particle events. The transient currents corresponding to each secondary particle are directly injected as PWL current sources into the netlist of the circuit to be evaluated, constructed based on the SPICE circuit model, for transient simulation. Based on the simulation results, it is determined whether a single-event upset soft error has occurred, and the soft error rate of the circuit is calculated based on statistical results. This step injects the physically-level precisely synthesized disturbance current into the circuit-level model and accurately captures the circuit's dynamic response through SPICE transient simulation. This achieves a complete mapping and quantitative evaluation from device physical effects to circuit functional errors, ultimately obtaining a reliable soft error rate index, providing a direct and efficient quantitative basis for radiation hardening design. Attached Figure Description

[0016] Figure 1 The flowchart of the method for rapid evaluation of single-event effects and soft error rate in hybrid simulation using an analytical function library is provided in the embodiments of the present invention. Figure 2A flowchart for constructing a transient pulse current function library provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of a priori sampling simulation provided in an embodiment of the present invention; Figure 4 A schematic diagram of a diffusion collection model in a three-dimensional Cartesian coordinate system provided in an embodiment of the present invention; Figure 5 A detailed flowchart of obtaining transient pulse current based on Geant4 simulation is provided for embodiments of the present invention; Figure 6 A flowchart for determining circuit-level soft errors provided in an embodiment of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0018] This invention provides a fast method for evaluating the soft error rate caused by single-event effects based on an analytical function library. See [link to relevant documentation]. Figure 1 The process includes the following steps S101 to S105.

[0019] S101, Construct a sensitive device model and a corresponding SPICE circuit model based on the process design kit, and use the SPICE circuit model to electrically calibrate the sensitive device model to obtain a calibration device model with accurate electrical characteristics. Specifically, in step S101, a sensitive device model and a corresponding SPICE circuit model are constructed based on the process design kit. The sensitive device model is then electrically calibrated using the SPICE circuit model to obtain a calibrated device model with accurate electrical characteristics. This includes the following steps S1011 to S1013.

[0020] S1011, based on the process design kit and manufacturing parameters of the target integrated circuit, constructs a sensitive device model in TCAD software and a corresponding SPICE circuit model in SPICE software; S1012, the key static electrical characteristics are simulated by constructing the sensitive device model and SPICE circuit model, and the simulation characteristic curve and electrical simulation characteristic curve are obtained accordingly. S1013, compare the electrical simulation characteristic curve and the simulation characteristic curve to adjust the parameters of the sensitive device model, and obtain a calibration device model with accurate electrical characteristics.

[0021] For example, device physical modeling and electrical calibration. Based on the process design kit (PDK) and manufacturing parameters of the target integrated circuit, an accurate three-dimensional physical model of the sensitive device is constructed in TCAD and SPICE software. By simulating its key static electrical characteristics, such as the Id-Vg transfer characteristic curve, and comparing the simulation results with those of the SPICE model based on the same PDK, parameter adjustments are made to complete the electrical calibration of the TCAD device model, ensuring its physical accuracy.

[0022] S102, based on the multidimensional key irradiation variables that affect transient current, performs TCAD transient simulation on the corrected device model under dense sampling conditions, and performs parameterized fitting on the transient current waveform obtained from the simulation based on physical mechanism, so as to construct a transient current response function library that correlates irradiation conditions and transient current parameters. Specifically, in step S102, based on the multidimensional key irradiation variables that affect the transient current, the corrected device model is subjected to TCAD transient simulation under dense sampling conditions, and the transient current waveform obtained from the simulation is subjected to parameterized fitting based on physical mechanisms to construct a transient current response function library that correlates irradiation conditions with transient current parameters, including the following steps S1021 to S1024.

[0023] S1021 defines a multidimensional key irradiation variable that affects transient current, and defines a sampling space based on the multidimensional key irradiation variable; wherein, the multidimensional key irradiation variable includes: particle incident position, linear energy transfer value and incident angle; S1022, the sampling density in the sampling space is determined based on prior simulation, and the transient response current waveform of the calibration device model under the irradiation conditions corresponding to the sampling point is obtained by TCAD simulation at each sampling point; S1023, The transient response current waveform is parameterized and fitted using a device response model based on physical mechanisms to obtain a current parameter vector with device physical meaning; here, the device response model based on physical mechanisms is expressed as: (1.1) in, Indicates the carrier drift velocity; Indicates linear energy transfer value; This represents the minimum distance a charge carrier must travel from its generation to its collection. Indicates carrier lifetime; Represents the bipolar diffusion coefficient of charge carriers; It represents the projected distance of the line segment AP formed by the particle incident point A and the carrier collection point P on the particle incident trajectory; This represents the projected distance of the line segment BP formed from the exit point B to the carrier collection point P on the particle's incident trajectory. Represents the Gaussian error function; Represents the carrier collection point P Axis coordinate values; Represents the carrier collection point P Axis coordinate values; This represents the time variable calculated from the point of incidence A.

[0024] S1024. Establish the mapping relationship between multidimensional key irradiation variables and corresponding current parameter vectors to obtain a transient current response function library that associates multidimensional key irradiation variables with transient current parameters.

[0025] For example, a library of transient current response functions can be constructed. (e.g.) Figure 2 As shown, Figure 2 The flowchart for constructing the transient pulse current function library is as follows: (1) Sampling space definition and prior simulation: The key irradiation variables affecting transient current are systematically defined, including particle incident position (X, Y, Z), linear energy transfer value (LET), and incident direction ( (etc.) Spatial signals characterizing the effects of irradiation, such as the collected charge of sensitive devices, are obtained through preliminary "a priori simulations." Based on the Nyquist-Shannon sampling theorem, a reasonable sampling density for sufficient sampling in the multidimensional variable space is determined, as illustrated in the sampling diagram. Figure 3 As shown, Figure 3 This is a schematic diagram of a priori sampling simulation.

[0026] (2) Dense sampling and TCAD simulation: Under a determined sampling density, the transient response current of the device under each set of sampling irradiation conditions is simulated in TCAD to obtain a series of "irradiation condition excitation-transient response current" mapping data pairs.

[0027] (3) Current parameter fitting and function library construction: The transient response current is parameterized and fitted based on the device response model with a well-defined physical mechanism. Possible models include, but are not limited to, the integral-based diffusion-collection model, whose Green's function form is shown in equation (1.1). The geometric relationship of this model in the three-dimensional coordinate system is as follows: Figure 4 As shown, Figure 4 This is a diffusion collection model in a three-dimensional Cartesian coordinate system.

[0028] By fitting, each transient response current is transformed into a set of transient current parameter vectors with device physical meaning, such as carrier diffusion coefficient D, carrier lifetime τ, and drift velocity. wait.

[0029] Finally, the irradiation condition variables of all sampling points are associated with the corresponding transient current parameter vectors to construct an "irradiation condition-transient current parameter" mapping database, which is the transient current response function library described in this invention.

[0030] S103, based on the actual layout of the integrated circuit, performs three-dimensional particle transport and nuclear reaction simulation under real radiation environment in Geant4 to obtain the set of secondary particles of a single simulated event acting on the correction device model. Specifically, in step S103, based on the actual layout of the integrated circuit, a three-dimensional particle transport and nuclear reaction simulation under a real radiation environment is performed in Geant4 to obtain a set of secondary particles that act on the correction device model by a single simulated event, including the following steps S1031 to S1033.

[0031] S1031, based on the actual layout of the integrated circuit, defines the geometry and material composition of the devices involved in the simulation, and uses the defined geometry and material data to configure the particle transport simulation environment in Geant4; S1032 performs Monte Carlo simulation based on the particle energy spectrum of the target radiation environment to track particle incidence, nuclear reaction, and secondary particle transport processes within the calibration device model. S1033, extract and record the set of secondary particles generated within the sensitive volume of the calibration device model in each simulation event.

[0032] For example, Geant4 radiation environment simulation and secondary particle information extraction. Based on the actual layout of integrated circuits, i.e., GDS files, three-dimensional device and circuit models including materials and geometry are constructed in Geant4. Figure 5 As shown, Figure 5 This document presents a detailed flowchart for obtaining transient pulse currents based on Geant4 simulation. It simulates a real radiation environment, including particle incidence, nuclear reactions, and the generation and transport of secondary particles such as alpha particles and heavy ions under the atmospheric neutron spectrum. It records and outputs detailed information on all secondary particles acting on the sensitive device during each valid event, including their type, energy, three-dimensional position, and trajectory length.

[0033] S104: Using the secondary particle set as the query condition, perform query and interpolation in the transient current response function library to obtain the corresponding transient current; Specifically, in step S104, the secondary particle set is used as a query condition to perform a query and interpolation in the transient current response function library to obtain the corresponding transient current, including: (1) Determine the number of secondary particles in the secondary particle set; If the number of secondary particles is equal to 1, the information of the secondary particles is used as the query condition, and matching and multidimensional interpolation are performed in the transient current response function library to obtain the corresponding transient current parameter vector, and the reconstructed transient current waveform is used as the corresponding transient current. If the number of secondary particles is greater than 1, then for each secondary particle, the information of the secondary particle is used as a query condition to perform matching and multidimensional interpolation in the transient current response function library to obtain the transient current parameter vector corresponding to each particle; the transient current waveform corresponding to each particle acting alone is reconstructed; all the reconstructed transient current waveforms are linearly superimposed, and the superimposed total waveform is taken as the corresponding transient current.

[0034] For example, a fast transient current synthesis based on a function library. Fast current synthesis: Information about each secondary particle output by Geant4 is used as a query condition, matched and interpolated in the constructed response function library to quickly obtain the transient current parameters generated when the particle acts alone, and then reconstruct its current waveform. For cases where a single event generates multiple secondary particles, such as a neutron nucleus reaction, the linear superposition algorithm shown in formula (1.2) is used to synthesize the total transient current response of the sensitive node under this event: (1.2) S105 uses the transient currents corresponding to each secondary particle as PWL current sources. Transient simulation is performed by directly injecting the circuit netlist of the circuit to be evaluated, which is constructed based on the SPICE circuit model. The simulation results are used to determine whether a single-event upset soft error has occurred, and the soft error rate of the circuit is calculated based on the statistical results.

[0035] Specifically, in step S105, it is determined whether a single-event upset soft error has occurred based on the simulation results, including the following steps S1051 to S1053.

[0036] S1051 monitors the voltage response of key nodes in the netlist of the circuit to be evaluated during SPICE transient simulation. S1052 compares the voltage response with a preset logic threshold; S1053 If the voltage response exceeds the logic threshold and causes the logic state to flip, a single-event upset soft error is determined to have occurred.

[0037] Specifically, in step S105, determining whether a single-event upset soft error has occurred based on the simulation results includes: (1) Obtain the static noise margin of the key nodes in the network table to be evaluated; wherein, the static noise margin is the maximum transient current peak or charge that the key nodes in the network table to be evaluated can withstand without logic flipping. (2) Compare the peak value or charge of the transient current with the static noise margin; (3) If the peak value of the transient current or the amount of charge exceeds the static noise tolerance, a single-event upset soft error is determined to have occurred. (4) If the peak value of the transient current or the amount of charge does not exceed the static noise tolerance, it is determined that no single-event upset soft error has occurred.

[0038] Specifically, in step S105, the formula for calculating the soft error rate of the circuit based on statistical results is as follows: (1.3) in, Indicates the number of soft errors; This indicates the total betting volume.

[0039] For example, circuit-level soft error detection: the synthesized total transient current It is injected into the circuit netlist to be evaluated built in SPICE software in two forms.

[0040] The first method: synthesizing from Geant4 As a PWL current source Direct injection into the SPICE netlist. This method can completely reproduce the fully coupled interaction between transient pulses and the dynamic characteristics of the circuit, reflecting the circuit's response to transient pulses under real-world conditions.

[0041] The second method involves artificially injecting transient pulse current into the circuit and, combined with theoretical analysis, extracting a static noise margin that characterizes the maximum pulse current intensity that sensitive nodes in the circuit can withstand. The peak value or charge quantity of the injected current is compared to this margin; if it exceeds the margin, an error is considered to have occurred. A comparison of the two methods is shown below. Figure 6 As shown.

[0042] Soft error section The calculation formula is shown in formula (1.3), molecule To determine the number of soft errors that occurred; denominator The total flux of particles, i.e., the total number of neutrons incident on a unit area of ​​the target device, is expressed in units of... The soft error cross section represents the sensitivity of an electronic device to single-event effects. The soft error cross section of the device can be used to calculate the soft error rate SER of the device in a real environment. The mathematical connection between the two is expressed in formula (1.4).

[0043] (1.4) The particle flux rate is the number of particles injected per hour per unit area of ​​the target device, in real-world conditions. ; It is usually characterized by FIT (Failure In Time), which refers to the device's performance in an irradiated environment. The number of soft errors that occurred within an hour.

[0044] The various embodiments described in this specification are presented in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. Each embodiment focuses on its differences from other embodiments. All or part of this invention can be used in numerous general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, mobile communication terminals, multiprocessor systems, microprocessor-based systems, programmable electronic devices, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices, etc.

[0045] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the present invention.

Claims

1. A fast evaluation method for soft error rate caused by single-event effects based on an analytical function library, characterized in that, include: Based on the process design kit, a sensitive device model and a corresponding SPICE circuit model are constructed. The SPICE circuit model is then used to perform electrical calibration on the sensitive device model to obtain a calibration device model with accurate electrical characteristics. Based on the multidimensional key irradiation variables that affect transient current, TCAD transient simulation under dense sampling conditions is performed on the corrected device model, and the transient current waveform obtained from the simulation is parametrically fitted based on physical mechanisms to construct a transient current response function library that correlates irradiation conditions with transient current parameters. Based on the actual layout of the integrated circuit, three-dimensional particle transport and nuclear reaction simulations under real radiation environment were performed in Geant4 to obtain the set of secondary particles of the correction device model under a single simulated event. Using the set of secondary particles as a query condition, the transient current response function library is queried and interpolated to obtain the corresponding transient current; The transient currents corresponding to each secondary particle are used as PWL current sources. Transient simulation is performed by directly injecting the circuit netlist constructed based on the SPICE circuit model into the circuit to be evaluated. Based on the simulation results, it is determined whether a single-event upset soft error has occurred, and the soft error rate of the circuit is calculated based on the statistical results.

2. The method for rapid evaluation of soft error rate caused by single-event effects based on analytical function library according to claim 1, characterized in that, The process of constructing a sensitive device model and a corresponding SPICE circuit model based on a process design kit, and then using the SPICE circuit model to perform electrical calibration on the sensitive device model to obtain a calibrated device model with accurate electrical characteristics includes: Based on the process design kit and manufacturing parameters of the target integrated circuit, a sensitive device model is built in TCAD software, and a corresponding SPICE circuit model is built in SPICE software. The key static electrical characteristics of the constructed sensitive device model and the SPICE circuit model are simulated to obtain the corresponding simulation characteristic curves and electrical simulation characteristic curves. By comparing the electrical simulation characteristic curve with the simulation characteristic curve, the parameters of the sensitive device model are adjusted to obtain a calibration device model with accurate electrical characteristics.

3. The method for rapid evaluation of soft error rate caused by single-event effects based on analytical function library according to claim 1, characterized in that, Based on the multidimensional key irradiation variables affecting transient current, the corrected device model undergoes TCAD transient simulation under dense sampling conditions, and the simulated transient current waveform is parametrically fitted based on physical mechanisms to construct a transient current response function library that correlates irradiation conditions with transient current parameters, including: A multidimensional key irradiation variable affecting transient current is defined, and a sampling space is defined based on the multidimensional key irradiation variable; wherein, the multidimensional key irradiation variable includes: particle incident position, linear energy transfer value and incident angle; The sampling density in the sampling space is determined based on prior simulation, and the transient response current waveform of the correction device model under the irradiation conditions corresponding to the sampling point is obtained by TCAD simulation at each sampling point. The transient response current waveform is parametrically fitted using a device response model based on physical mechanisms to obtain a current parameter vector with device physical meaning; Establish the mapping relationship between the multidimensional key irradiation variables and the corresponding current parameter vectors to obtain a transient current response function library that associates the multidimensional key irradiation variables with transient current parameters.

4. The method for rapid evaluation of soft error rate caused by single-event effects based on analytical function library according to claim 3, characterized in that, The device response model based on the physical mechanism is expressed as follows: ； in, Indicates the carrier drift velocity; Indicates linear energy transfer value; This represents the minimum distance a charge carrier must travel from its generation to its collection. Indicates carrier lifetime; Represents the bipolar diffusion coefficient of charge carriers; It represents the projected distance of the line segment AP formed by the particle incident point A and the carrier collection point P on the particle incident trajectory; This represents the projected distance of the line segment BP formed from the exit point B to the carrier collection point P on the particle's incident trajectory. Represents the Gaussian error function; Represents the carrier collection point P Axis coordinate values; Represents the carrier collection point P Axis coordinate values; This represents the time variable calculated from the point of incidence A.

5. The method for rapid evaluation of soft error rate caused by single-event effects based on analytical function library according to claim 1, characterized in that, The actual layout based on the integrated circuit is used in Geant4 to simulate three-dimensional particle transport and nuclear reactions under a real radiation environment, resulting in a set of secondary particles acting on the correction device model under a single simulated event, including: Based on the actual layout of the integrated circuit, the geometry and material composition of the devices involved in the simulation are defined, and the particle transport simulation environment is configured in Geant4 using the defined geometry and material data; Based on the particle energy spectrum of the target radiation environment, a Monte Carlo simulation is performed to track the particle incident, nuclear reaction, and secondary particle transport processes within the model of the correction device. Extract and record the set of secondary particles generated within the sensitive volume of the calibration device model during each simulation event.

6. The method for rapid evaluation of soft error rate caused by single-event effects based on analytical function library according to claim 1, characterized in that, The step of using the secondary particle set as a query condition to perform querying and interpolation in the transient current response function library to obtain the corresponding transient current includes: Determine the number of secondary particles in the set of secondary particles; If the number of secondary particles is equal to 1, then the information of the secondary particles is used as a query condition to perform matching and multidimensional interpolation in the transient current response function library to obtain the corresponding transient current parameter vector, and the reconstructed transient current waveform is used as the corresponding transient current. If the number of secondary particles is greater than 1, then for each secondary particle, the information of the secondary particle is used as a query condition to perform matching and multidimensional interpolation in the transient current response function library to obtain the transient current parameter vector corresponding to each particle; the corresponding transient current waveform under the individual action of each particle is reconstructed; all the reconstructed transient current waveforms are linearly superimposed, and the superimposed total waveform is used as the corresponding transient current.

7. The method for rapid evaluation of soft error rate caused by single-event effects based on analytical function library according to claim 1, characterized in that, The method for determining whether a single-event upset soft error has occurred based on simulation results includes: During the SPICE transient simulation, the voltage response of key nodes in the netlist of the circuit to be evaluated is monitored; The voltage response is compared with a preset logic threshold; If the voltage response exceeds the logic threshold and causes the logic state to flip, a single-event upset soft error is determined to have occurred.

8. The method for rapid evaluation of soft error rate caused by single-event effects based on analytical function library according to claim 1, characterized in that, The method for determining whether a single-event upset soft error has occurred based on simulation results includes: Obtain the static noise margin of the critical nodes in the network table of the circuit to be evaluated; wherein, the static noise margin is the maximum transient current peak or charge that the critical nodes in the network table of the circuit to be evaluated can withstand without logic flipping; The peak value or charge amount of the transient current is compared with the static noise tolerance; If the peak value or charge of the transient current exceeds the static noise tolerance, a single-event upset soft error is determined to have occurred. If the peak value or charge of the transient current does not exceed the static noise tolerance, then it is determined that no single-event upset soft error has occurred.

9. The method for rapid evaluation of soft error rate caused by single-event effects based on analytical function library according to claim 1, characterized in that, The formula for calculating the soft error rate of the circuit based on statistical results is as follows: ； in, Indicates the number of soft errors; This indicates the total betting volume.

Images