Voronoi Diagram Pattern Matching for IC Layout Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Design Rule Checking (DRC) tools for integrated circuit design suffer from high false positive and false negative rates and require specialized programming skills, while image processing tools lack saliency, making it difficult to identify salient design patterns effectively.
Innovation Solution
The method employs Voronoi diagrams to compute edge regions for both reference and search spaces, allowing users to select salient regions and construct search patterns, which are then used to highlight matching patterns in the search space, simplifying the identification of salient design shapes without requiring specialized knowledge.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If Design Rule Checking (DRC) tools are used for pattern matching, then false positive and false negative rates are reduced, but specialized programming skills are required
Solution Approach 1:
The patent introduces an image processing intermediary layer between the user and the DRC tool. This intermediary automatically converts user-selected visual features into DRC-readable pattern definitions, allowing users to interact with simple image processing tools while the system handles the complex DRC pattern encoding in the background, thus resolving the contradiction between high accuracy and ease of use
Solution Approach 2:
The system creates a copy of the reference configuration from the search space, applies image processing to identify salient features in this copy, and then uses these features to define search patterns. This copying approach allows the complex pattern definition process to occur on a duplicate rather than the original design, maintaining accuracy while simplifying user interaction
2Ease of operation
If image processing tools are used for pattern matching, then ease of operation is improved, but false positive and false negative rates increase due to lack of saliency
Solution Approach 1:
The system performs preliminary action by automatically computing Voronoi diagrams and identifying salient features from the user-selected image processing results before the actual pattern matching occurs. This preliminary processing ensures that only the most relevant geometric features are used to define search patterns, thereby maintaining high accuracy while keeping the user interface simple
Solution Approach 2:
The patent replaces the manual mechanical process of defining complex DRC patterns with an automated computational system that uses Voronoi diagrams and image processing. This substitution eliminates the need for users to manually encode pattern features while maintaining the precision of DRC-level analysis through automated feature extraction
3Measurement precision
If complex DRC pattern definitions are used, then pattern matching accuracy is improved, but device complexity increases
Solution Approach 1:
The patent segments the complex pattern definition process into distinct automated stages: Voronoi diagram computation, edge region identification, salient feature extraction, and search pattern generation. By dividing this complex process into manageable automated segments, the system achieves high pattern matching accuracy while keeping the overall system complexity manageable through modular design
Data Source
AI summary
A method for implementing pattern matching of integrated circuit features includes computing Voronoi edge regions for both a reference configuration and a search space of an integrated circuit design to be searched and presenting the computed Voronoi edge regions of the reference configuration to a user; receiving one or more selected bisectors of the Voronoi computed reference configuration from the user, indicative of user identified salient regions of design shapes and/or corners to be searched, so as to define one or more search elements, wherein a search element comprises a given bisector and a pair of Voronoi edge regions bounded thereby; constructing a search pattern from the one or more search elements defined from the reference configuration; examining the search space for matching sequences with respect to the search pattern; and highlighting resulting matching patterns in the search space for the user.