Design And Implementation Of PD Radar Signal Processor Verification Platform Based On UVM

This project is based on a national PD radar signal processor project.In this paper,the function verification of PD radar signal processor module is researched deeply,and the verification target of PD radar signal processor module is determined.The algorithm based on genetic algorithm and the functional coverage model and the PD radar based on SVA are designed to verify the assertion module,finally building the UVM random verification platform for the PD radar signal processor module.PD radar signal processor module belongs to the algorithm function module,the correctness of the work is reflected in the correct implementation of the algorithm,so the verification goal of this paper is to build a verification platform to ensure the correct implementation of the module algorithm.To achieve this goal,a sufficient number of validation vectors are used to motivate the module and the results of the module are compared with the correct results after the reference model is processed.In order to speed up the convergence rate and improve the verification efficiency,a validation vector is generated using a coverage-based driving method based on genetic algorithm.The vector generation method combines the input data characteristics of the PD radar signal processor module,performs mathematical and probabilistic correlation analysis on the genetic operators,and finally selects the proportional selection operator,the uniform crossover operator and the binary mutation operator and applies it to the verification Generate vectors in the platform.Second,the PD radar signal processor module internal finite state machine and logic control signal jump state is complex,the timing of the module control signal to meet the design specifications,in order to real-time monitoring signal changes,improve verification of observability,fast positioning error and shorten the verification Cycle,to the PD radar signal processor verification process to increase the assertion detection points.Based on the above verification scheme,a hierarchical UVM verification platform is designed and implemented to complete the verification of PD radar signal processor module.Coverage is a measure of the completeness of functional verification and has a very important role in functional verification.According to the difference of function coverage,this topic is divided into two kinds of statistical methods: incentive coverage and assertion coverage.Based on the final coverage statistics,the verification work is summed up.Simulation results using the VCS simulation tool show that the verification vector generation technique using genetic algorithm is faster than the traditional pure random verification vector generation technique and increases by about 16.7%.All 35 assertion verification modules are put into use and find one error;functional coverage model coverage of more than 98%.