Data Characteristics Deconstruction Method For Signal Integrity Simulation Results Credibility Evaluation

In recent years, with the maturity of the theory of computational electromagnetics and the possible of a large number of the commercial electromagnetic compatibility softwares, early in the design of high speed PCB board the systematic analysis of the signal integrity (SI) simulation can be achieved, but whether the results obtained by simulation software reflect the objective real state, we need a complete and effective set of evaluation methods to verify. Therefore, it is urgent that we put forward the effective method of the credibility quantitative assessment of the SI simulation results.In this paper, the "data characteristics deconstruction method" is first raised to simulate the visual assessment mechanism to quantificationally assess the credibility of the SI simulation results and mirror the SI simulation problems'characteristics in the domestic.First, the typical problems of the signal integrity such as the reflection, crosstalk, the noise of the differential microstrip line and via are analysed to get the typical features of the SI simulation issues, these features are summarized to establish the SI data characteristics library. Second, the characteristics extracting method such as the Fourier transformation, wavelets, Prony transformation, relevance theory and statistical methods are raised besed on the characteristics of the above features of the library. Then, the extracted SI features are analyzed to build the different evaluation indicators. Finally, the "AHP" are used to determine the weight of the evaluation indicators so that the global credibility is synthesized to complete the construction of the method.The data characteristics deconstruction method is applied to the typical SI problems such as the S parameters and the ringing problem to assess the credibility of the simulation results.The results between the assessment method and the assessment of the researchers in the related fields are consistent, so the data characteristics deconstructing method of the quantitative assessment of the SI simulation results'credibility is correct and effective.