New methodology for digital design properties extraction from simulation traces

Abstract

This paper introduces a new methodology for digital design properties extraction from simulation traces. The innovated methodology is based on a new data mining technique guided with static analysis of the intended design. The mining engine of the proposed methodology is based on innovated Breadth-First Decision Tree (BF-DT) search algorithm. The data structure of each node in the decision tree is handled to well present sub-space of the input simulation traces data space. Besides, new features are added to BF-DT to enhance its performance in both output sequential assertions and time of search. A new static analysis technique is innovated to extract all the combinational and sequential data dependencies between the digital design signals. The mining engine is guided with these data dependencies to extract complete combinational and sequential design properties relating signals desired to extract properties for and their cone of interest signals. The contributed mining technique has been tested for bit-level designs with different sizes. The design properties generated from the mining engine completely match with all design properties covered in the input simulation traces. Plus, the generated properties are at the highest possible level of abstraction leading to the best coverage for the input data space. The simulation results show that the proposed methodology has proven superior efficiency in extracting bit-level assertions of digital design in a feasible time. The next challenge is to include word-level assertions as well.