| Large scale integrated circuits receive cosmic radiation such as ions,protons and gamma rays in space.These high-energy particles will accumulate electric charge in a short time,which reduces the radiation resistance of spacecraft.Compared with space carrying experiments and ground simulation experiments with long period and high cost,computer simulation technology built a new way for evaluating the spacecraft.Computer uses fault injection method to simulate Single Event Effect,which can verify the reliability of the target system in time,quickly and at low cost.It is of great significance to evaluate the reliability of the circuit.Aiming at the high reliability requirements of aerospace circuits,this paper creatively proposes a fast simulation method of Single Event Effect to solve the problem of long cycle and high overhead in reliability evaluation.Through the sensitive path search and extraction of large-scale integrated circuits,fault injection sets with different sensitive levels are established.Constructed the Single Event Effect simulation method based on the fault injection set,fault injection model and fault injection technology.The main research results are as follows:(1)A sensitive path search and extraction method for large-scale integrated circuits is realized.Using the static timing analysis method to search sensitive path,all logic gates on the path will be extracted.By analyzing the characteristics of SPICE netlist and extracting all sensitive nodes of the target circuit,the fault injection sets with different sensitivity levels is built.On this basis,the extracted sensitive path is verified by Single Event Effect simulation.The simulation results show that the path with the most stringent timing margin is more vulnerable to single event effect,resulting in errors in the internal registers of the circuit.(2)A Single Event Effect simulation method for large-scale integrated circuits is realized.The simulation method is composed of fault injection model,fault injection technology and spice level simulation method.The transient response of multi pulse integrated circuit is simulated by the method of particle injection current,which is suitable for the simulation of integrated circuit.According to the fault injection model and fault injection technology,the SPICE simulation model of the target circuit is built.By using Finesim tool,a fast simulation method for large-scale integrated circuits is realized.Finally,a fast simulation evaluation method is explored to simplify the fault injection set and reduce the number of fault injection.This method can test the anti-Single Event Effect ability of the system in a limited time by injecting the fault in the more sensitive fault set.(3)The SEE simulation analysis of LEON2 processor based on SPARC V8 architecture is completed.This paper mainly analyzes the Integer Unit,Register,Cache,AHB Controller and Memory Control module in the system.The results show that the SEE sensitivity of different modules in the processor system is different.The sensitivity of Integer Unit module is the highest,followed by Register,Cache and Memory Control module,and the sensitivity of AHB controller is the lowest.The simulation results show that the fault injection conforms to the propagation law of Single Event Effect,and the SPICE simulation method can effectively evaluate the reliability of integrated circuits.(4)The SEE simulation analysis of CV32E40 P processor based on RISC-V instruction set architecture is completed.It mainly completes the simulation and evaluation of the fourstage pipeline module in the processor.The simulation results show that among Instruction Fetch,Instruction Decode,Execute and Write Back stages,the single event effect sensitivity of Instruction Decode stage is the highest,the sensitivity of Instruction Fetch stage is the second,the next is Execute stage,and the sensitivity of Write Back stage is the lowest.Finally,the reliability rapid evaluation method is used to test the more sensitive path fault injection set.This method reduces the number of invalid injections and improves the evaluation efficiency of the system. |