Circuit original input end sensitivity calculation method based on probability model

Abstract

The invention discloses a circuit original input end sensitivity calculation method based on a probability model. The method comprises the steps of carrying out netlist analysis and correlation quantity initialization; judging the criticality Crtli, j of the jth input end of the ith node in the LC; pruning the circuit according to the result to obtain a circuit sensitive node corresponding to theapplied input vector, and allocating a sensitive value to the marked sensitive node; and giving the clustering number of sen based on kernel density estimation, clustering the sen by using a k-means algorithm, and outputting a clustering result according to the size sequence of the sensitivity values. Marking of a sensitive original input end under a given input vector is realized based on a pruning strategy of a shielding effect; the sensitivity level of each original input end of the circuit is quantized based on a reverse depth-first search strategy by means of an SCA algorithm according tothe type and topological position information of the nodes; and in combination with a clustering algorithm, effective identification of the sensitive original input end in the circuit is realized through continuous iteration.

Description

technical field [0001] The invention relates to the sensitivity calculation of the original input terminal of an integrated circuit, specifically a calculation method based on the combination of a pruning strategy and a shielding effect measurement. Background technique [0002] The continuous shrinking of the feature size of integrated circuits inevitably leads to the decrease of circuit reliability tolerance. Existing studies have shown that the reliability of circuits is affected by circuit input vectors, and the reliability differences under different input vectors sometimes even differ by several orders of magnitude. For this reason, in the process of circuit design, it is necessary to understand the boundary of circuit reliability and its corresponding input vector, so as to improve the reliability level of the circuit purposefully. Carrying out the analysis of the sensitivity of the original input terminal of the circuit is an effective method to achieve the above go...

AI Technical Summary

Problems solved by technology

However, when the above method is faced with a circuit with a complex structure, it often has a large calculation overhead and the self-learning ability of the algorithm needs to be improved to further improve the calculation accuracy

Images