Similar wiring method for target layout based on original layout wiring information

Abstract

The invention discloses a method for conducting similarity wiring on a target layout based on original layout wiring information, and belongs to the field of layout wiring design. The method comprises the following steps: firstly, acquiring routing information of an original layout, layer information of a target layout and wiring constraint, and extracting a topological structure vector and a geometric constraint vector of original routing; then, a line exploration similar wiring algorithm is called to be combined with the vector optimization exploration direction, if wiring succeeds, a result is directly drawn, if wiring fails, a wiring diagram with a similarity gradient is initialized based on the vector, iterative search is conducted through an A * search algorithm introducing a similarity evaluation item, and a result is obtained; and obtaining a wiring solution with the maximum similarity with the original wiring, and drawing the wiring solution to a target layout. According to the invention, by multiplexing the routing information of the original layout, the matching of the wiring of the target layout and the original design in topological and geometrical characteristics is realized, and meanwhile, the process constraint of the target layout is adapted, so that the problems of poor circuit performance consistency and low design efficiency in the prior art are solved.

Description

Technical Field

[0001] This invention relates to the field of layout routing design, and more specifically to a method for routing a target layout with similarity based on the original layout routing information. Background Technology

[0002] Current mainstream automatic routing methods focus on improving routing success rate, meeting integrated circuit design constraints, optimizing timing performance, reducing signal crosstalk, and ensuring signal integrity. These goals are typically achieved through maze algorithms, line exploration algorithms, or pattern matching algorithms. While maze algorithms offer high routing success rates, they are relatively slow. Line exploration and pattern matching algorithms, while faster, suffer from lower routing success rates. A clear trade-off exists between these two approaches in the "speed-routing success rate" performance dimension.

[0003] These automated routing methods significantly reduce the time engineers spend on manual routing. In large-scale integrated circuit design scenarios, where manual routing involves an enormous workload (covering a massive number of devices and circuit connections), automated routing has become an inevitable choice in engineering practice. However, in business scenarios involving cross-process layout migration of analog chips and rerouting after local device adjustments in the layout, using mainstream fully automated routing methods cannot utilize the information from the original routing to extract user habits, resulting in routing results that differ significantly from the user's expected original results.

[0004] In the above scenarios, the core requirement for routing is "rerouting while preserving the original wiring pattern," rather than simply achieving conductivity between devices. Existing automatic routing methods fail to address this requirement and cannot meet actual design needs. Therefore, engineers still need to manually complete the routing operations in EDA (Electronic Design Automation) software, which not only consumes a significant amount of time but may also introduce consistency errors due to manual operation, affecting design efficiency and layout reliability. Summary of the Invention

[0005] To address the shortcomings of existing technologies, the purpose of this invention is to provide a method for routing a target layout with similarity based on the original layout routing information. By reusing the original layout routing information, the routing of the target layout is matched with the original design in terms of topology and geometric features. At the same time, it adapts to the process constraints of the target layout, solves the problems of poor circuit performance consistency and low design efficiency in existing technologies, and improves the reliability and efficiency of layout routing.

[0006] To achieve the above objectives, the embodiments of this invention provide the following technical solutions:

[0007] This application provides a method for performing similarity routing on a target layout based on the routing information of the original layout, including the following steps: S101, obtaining the routing information of the original layout, the routing information including the position, physical size, layer, and network information of lines and vias; S102, obtaining the layer information of the target layout and establishing a correspondence with the layer information of the original layout, and reading the routing constraint information on each layer of the target layout; S103, extracting the topology vector and geometric constraint vector of the original routing based on the routing information of the original layout; S104, calling the similar routing algorithm to explore routing, using the topology vector and geometric constraint vector to optimize the exploration direction during the exploration process; if the routing is successful, proceed to step S107, otherwise proceed to step S105; S105, initializing a similar routing graph based on the topology vector and the geometric constraint vector; S106, iteratively executing the A* search algorithm on the similar routing graph to search for routing paths until obtaining the routing solution with the highest similarity; S107, drawing the routing solution on the target layout.

[0008] Furthermore, in step S102, the wiring constraint information includes minimum line width and minimum line spacing.

[0009] Further, step S105 includes: S1051, using the center of the topological structure vector and the geometric constraint vector as a reference, causing the similarity gradient of the grids in the similar wiring diagram to spread outward according to a Gaussian distribution; S1052, assigning a higher candidate priority to the grid where the center is located according to the similarity gradient, thereby completing the initialization of the similar wiring diagram.

[0010] Further, step S106 includes: S1061, introducing a similarity evaluation term calculated based on the topological structure vector and geometric constraint vector into the heuristic function of the A* search algorithm, so as to guide the search path to preferentially approach the center of the topological structure vector and geometric constraint vector and automatically bypass obstacles.

[0011] Furthermore, in step S104, the line exploration similarity routing algorithm is a meshless routing algorithm. In the process of selecting escape points, the meshless routing algorithm combines the topological structure vector and the geometric constraint vector to control the direction of the escape points.

[0012] Further, in step S106, the similarity of the wiring results is evaluated using a similarity measurement formula, which is: ; ; ;in, For topological similarity, For geometric similarity, Indicates wiring reference, Indicates similar routing results. Indicates the final similarity. Indicates the intermediate similarity. The weights for topological similarity. The weights for geometric similarity.

[0013] Furthermore, the topological similarity The calculation formula is: ;in, Topology sequence for wiring reference Topology sequence of wiring results The Levinstein distance between them For the length of the wiring reference topology sequence, The length of the topology sequence of the wiring result.

[0014] Furthermore, the geometric similarity The calculation formula is: ; ;in, For length similarity, For the similarity of the included angle, The weights for length similarity. The weights for the similarity of the included angles.

[0015] Furthermore, the length similarity The calculation formula is: ;in, A vector representing the segment length ratio for routing reference. This is the vector representing the length ratio of the line segments in the wiring result.

[0016] Furthermore, the included angle similarity The calculation formula is: ; ;in, The vector representing the angle between any two line segments in the routing reference. This is the vector of the angle between any two line segments in the wiring result. Let θ be the minimum length among θ and ϕ.

[0017] The beneficial effects of this invention are as follows: by reusing the original layout routing information, the target layout routing and the original design are matched in terms of topology and geometric features. At the same time, it adapts to the process constraints of the target layout, solves the problems of poor circuit performance consistency and low design efficiency in the prior art, and improves the reliability and efficiency of layout routing. Attached Figure Description

[0018] Figure 1This application provides a flowchart illustrating a method for routing a target layout with similarity based on the original layout routing information. Detailed Implementation

[0019] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the scope of the present invention.

[0020] In this invention, the terms "system" and "network" are used interchangeably. "Multiple" refers to two or more; therefore, in this invention, "multiple" can also be understood as "at least two." "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. Additionally, the character " / ", unless otherwise specified, generally indicates that the preceding and following related objects have an "or" relationship. Furthermore, it should be understood that in the description of this invention, terms such as "first" and "second" are used only for descriptive purposes and should not be construed as indicating or implying relative importance or order.

[0021] When routing existing layouts, the target layout is often processed independently, completely ignoring the routing design experience that has been verified in the original layout. This not only leads to significant differences between the routing results and the original design in terms of topology and geometry, but also causes circuit performance to deviate from expectations in scenarios such as analog circuits that are sensitive to routing consistency.

[0022] like Figure 1 As shown, this application embodiment provides a method for performing similarity routing on a target layout based on the routing information of the original layout, including the following steps: S101, obtaining the routing information of the original layout, the routing information including the position, physical size, layer, and network information of lines and vias; S102, obtaining the layer information of the target layout and establishing a correspondence with the layer information of the original layout, and reading the routing constraint information on each layer of the target layout; S103, extracting the topology vector and geometric constraint vector of the original routing based on the routing information of the original layout; S104, calling the similar routing algorithm to explore routing, using the topology vector and geometric constraint vector to optimize the exploration direction during the exploration process; if the routing is successful, proceed to step S107, otherwise proceed to step S105; S105, initializing a similar routing graph based on the topology vector and the geometric constraint vector; S106, iteratively executing the A* search algorithm on the similar routing graph to search for routing paths until obtaining the routing solution with the highest similarity; S107, drawing the routing solution on the target layout.

[0023] In another possible embodiment, the EDA design file storing the original layout data is first read to obtain complete routing information such as the location, physical size, layer, and network of lines and vias. Next, the layer information of the target layout is obtained, and a correspondence is established between the layers of the target layout and the original layout according to the principle of functional matching. Simultaneously, the routing constraint information of each layer of the target layout is read. Then, the topology vector reflecting the routing connection relationship and the geometric constraint vector reflecting the routing length, angle, and other characteristics are extracted from the original routing information. Afterward, a similar routing algorithm is called, combining the aforementioned topology and geometric vectors to optimize the direction at each step of the exploration. If the routing is successfully completed, the result is directly drawn onto the target layout. If it fails, a similar routing graph is initialized based on the topology and geometric vectors, and the A* search algorithm is iteratively executed on this graph until the routing solution with the highest similarity to the original routing is obtained. Finally, this solution is drawn onto the target layout.

[0024] On the one hand, by reusing the routing information of the original layout, the routing of the target layout can be matched with the original design in terms of topological connection logic and geometric layout characteristics, ensuring the consistency of circuit performance. On the other hand, the layer constraints of the target layout are adapted synchronously during the routing process, which not only avoids the redundant work of redesign, but also ensures that the routing meets the process requirements of the target layout, greatly improving the efficiency and reliability of routing.

[0025] If specific process parameter limitations are not clearly defined, the routing results may not meet the process requirements such as minimum line width and minimum line spacing of the target layout, leading to repeated adjustments or even rerouting.

[0026] In this embodiment of the application, in step S102, the wiring constraint information includes minimum line width and minimum line spacing.

[0027] In another possible embodiment, after obtaining the layer information of the target layout and establishing a correspondence with the original layout, the routing constraint information of each layer of the target layout is read synchronously, which specifically includes the minimum allowable line width parameter and minimum line spacing parameter of each layer, and these parameters are used as process constraints in the subsequent routing process.

[0028] By clearly defining the minimum line width and minimum line spacing constraints for each layer of the target layout, the routing process is always based on these parameters, avoiding the problem of routing results not conforming to process specifications from the source, reducing the cost of subsequent adjustments, and also ensuring the electrical performance and reliability of the routing.

[0029] Existing routing search methods lack guidance based on the original routing characteristics during the search process, and the search direction is relatively random. This not only increases redundant steps in the search but also leads to a significant deviation between the final routing result and the topological and geometric characteristics of the original design.

[0030] In this embodiment of the application, step S105 includes: S1051, taking the center of the topological structure vector and the geometric constraint vector as a reference, making the similarity gradient of the grid in the similar wiring diagram diffuse outward according to a Gaussian distribution; S1052, according to the similarity gradient, assigning a higher candidate priority to the grid where the center is located, and completing the initialization of the similar wiring diagram.

[0031] In another possible embodiment, when routing fails and a similar routing graph needs to be initialized, the center of the topology vector and geometric constraint vector of the original routing is used as a reference. The similarity gradient of each grid in the similar routing graph is allowed to diffuse outward from the center according to the Gaussian distribution (e.g., the gradient decreases Gaussian with the distance from the center). Based on this gradient distribution, the grid where the center is located is given a higher candidate priority than other grids in other regions, thereby completing the initialization of the similar routing graph.

[0032] By constructing a similarity gradient based on the original routing features, the search process prioritizes moving closer to the core feature regions of the original routing. This reduces invalid search steps and ensures that the routing results maintain a higher degree of matching with the original design in terms of topological and geometric features, thereby improving the efficiency and similarity of routing.

[0033] The heuristic function of the existing A* search algorithm only focuses on the optimization of path length and does not take into account the characteristics of the original routing. This not only causes the search path to deviate from the topological and geometric features of the original design, but also makes it easy to get stuck in the obstacle area of ​​the target layout, increasing the risk of routing failure.

[0034] In this embodiment of the application, step S106 includes: S1061, introducing a similarity evaluation term calculated based on the topological structure vector and geometric constraint vector into the heuristic function of the A* search algorithm, so as to guide the search path to preferentially approach the center of the topological structure vector and geometric constraint vector and automatically bypass obstacles.

[0035] In another possible embodiment, when performing the A* search algorithm on a similar routing graph, a similarity evaluation term calculated based on the original routing topology vector and geometric constraint vector is added to the heuristic function of the A* search algorithm. Based on the similarity evaluation term, the search path is guided to preferentially approach the central region of the topology vector and geometric constraint vector. At the same time, when encountering obstacles in the target layout, the path direction is automatically adjusted to bypass the obstacles, and the search continues until a routing solution that meets the requirements is obtained.

[0036] Introducing a similarity evaluation term into the heuristic function not only encourages the search path to actively approach the core feature area of ​​the original routing, ensuring the similarity of the routing, but also automatically avoids obstacle areas during the search process, balancing the feasibility and efficiency of routing and reducing redundant search operations.

[0037] Traditional mesh routing algorithms are limited by the fixed layout of the mesh, resulting in poor flexibility and difficulty in quickly adapting to the topology and geometric features of the original routing. This not only leads to low routing efficiency but also makes it difficult to guarantee the similarity between the final result and the original design.

[0038] In this embodiment of the application, in step S104, the line exploration similarity routing algorithm is a meshless routing algorithm. In the process of selecting escape points, the meshless routing algorithm combines the topological structure vector and the geometric constraint vector to control the direction of the escape points.

[0039] In another possible embodiment, when calling the line exploration similarity routing algorithm for routing, by taking advantage of its meshless nature, the extension direction of each escape point is controlled by combining the topological structure vector and geometric constraint vector extracted from the original routing during the selection process, thereby quickly generating a routing structure that is similar to the original routing in both topological and geometric features.

[0040] The algorithm employs a meshless line exploration similar routing algorithm, which eliminates the limitations of grid layout. It also combines the topology and geometric vectors of the original routing to control the direction of the escape point, which can more flexibly match the characteristics of the original routing and quickly generate similar routing results, thus improving the efficiency and similarity of routing.

[0041] Existing methods for assessing cabling similarity lack a unified and flexible quantitative approach, making it impossible to balance the weights of topology and geometric features according to different design scenarios. This results in assessment results that do not match actual design requirements, making it difficult to effectively guide cabling optimization.

[0042] In this embodiment of the application, in step S106, the similarity of the wiring results is evaluated using a similarity measurement formula, which is: ; ; ;in, For topological similarity, For geometric similarity, Indicates wiring reference, Indicates similar routing results. Indicates the final similarity. Indicates the intermediate similarity. The weights for topological similarity. The weights for geometric similarity.

[0043] In another possible embodiment, when evaluating the similarity of the wiring results, the topology similarity is calculated first. and geometric similarity Then according to the preset weight ratio ( ), through formula The intermediate similarity values ​​are obtained, and finally, the formula is used. By constraining the results to the range of 0 to 1, the final wiring similarity is obtained.

[0044] By integrating topological similarity and geometric similarity into a unified quantitative result through a weighted formula, it also supports adjusting the weight ratio of the two according to different design scenarios, making the similarity assessment more in line with actual needs, and more accurately guiding the optimization of the routing process to ensure that the routing results meet design expectations.

[0045] Current topological similarity assessments do not consider the length differences of topological sequences and rely solely on simple matching degree calculations. This results in assessments that fail to accurately reflect the degree of similarity between topological sequences of different lengths, thus affecting the reliability of similarity assessments.

[0046] In this embodiment of the application, the topological similarity The calculation formula is: ;in, Topology sequence for wiring reference Topology sequence of wiring results The Levinstein distance between them For the length of the wiring reference topology sequence, The length of the topology sequence of the wiring result.

[0047] In another possible embodiment, when calculating topology similarity, the topology sequence of the wiring reference is first obtained. Topology sequence of wiring results and calculate and Lewinstein distance between Next, determine the maximum length of the topology sequence of the routing reference and the topology sequence of the routing result, denoted as . Finally, through the formula Obtain the topological similarity.

[0048] By introducing Levenstein distance and combining it with the topology sequence length for normalization, the interference of sequence length differences on the evaluation results is eliminated, making the calculation of topology similarity more accurate and able to more realistically reflect the degree of similarity between the wiring result and the original design in terms of topology connection.

[0049] In this embodiment of the application, the geometric similarity The calculation formula is: ; ;in, For length similarity, For the similarity of the included angle, The weights for length similarity. The weights for the similarity of the included angles.

[0050] Current geometric similarity assessments do not differentiate between the two core geometric features of length and included angle, but only perform a general fuzzy assessment. This fails to emphasize different geometric features according to design requirements, resulting in assessment results that cannot accurately match the needs of the design scenario.

[0051] In another possible embodiment, when calculating geometric similarity, length similarity is calculated first. Similarity with angle Then according to the preset weight ratio ( ), through formula Obtain the geometric similarity.

[0052] By splitting geometric similarity into length similarity and angle similarity, and supporting adjustments to their weight ratios, the evaluation of geometric similarity can be adapted to the emphasis on length or angle features in different design scenarios, thus improving the flexibility and accuracy of geometric similarity evaluation.

[0053] Existing length similarity assessments only focus on the absolute length of line segments, without considering the length ratio between line segments. This results in the assessment results failing to accurately reflect the geometric similarity between the wiring results and the original design.

[0054] In this embodiment of the application, the length similarity The calculation formula is: ;in, A vector representing the segment length ratio for routing reference. This is the vector representing the length ratio of the line segments in the wiring result.

[0055] In another possible embodiment, when calculating length similarity... First, obtain the segment length ratio vector of the routing reference. The vector proportional to the segment length of the wiring result Then calculate the dot product of the two vectors, then calculate the product of the magnitudes of the two vectors, and finally use the formula... Calculate the length similarity.

[0056] By calculating based on the line segment length ratio vector, the length similarity assessment focuses on the relative proportional relationship between line segments, which better meets the needs of geometric layout similarity in actual design and improves the accuracy of geometric similarity assessment.

[0057] Current methods for assessing the similarity of included angles do not consider the differences in the length of the included angle vectors. They directly calculate all included angles uniformly, which leads to deviations in the assessment results due to differences in vector lengths, and fails to accurately reflect the degree of similarity of the included angle features.

[0058] In this embodiment of the application, the included angle similarity The calculation formula is: ; ;in, The vector representing the angle between any two line segments in the routing reference. This is the vector of the angle between any two line segments in the wiring result. Let θ be the minimum length among θ and ϕ.

[0059] In another possible embodiment, when calculating the angle similarity, the angle vectors between each pair of line segments in the wiring reference are first obtained. The vector of the angle between each pair of line segments in the wiring result Determine the smaller of the two lengths, and then, based on the determined smaller of the two lengths, use the formula... Obtain the angle similarity.

[0060] By using a shorter included angle vector as the benchmark for matching calculations, the interference of vector length differences on the evaluation results is eliminated, making the calculation of included angle similarity more accurate and able to more realistically reflect the degree of similarity between the wiring result and the original design in terms of included angle features.

[0061] The optional embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the embodiments of the present invention are not limited to the specific details in the above embodiments. Within the scope of the technical concept of the embodiments of the present invention, various simple modifications can be made to the technical solutions of the embodiments of the present invention, and these simple modifications all fall within the protection scope of the embodiments of the present invention.

[0062] It should also be noted that the various specific technical features described in the above embodiments can be combined in any suitable manner without contradiction. To avoid unnecessary repetition, the embodiments of the present invention will not describe the various possible combinations separately.

[0063] Furthermore, various different implementations of the present invention can be combined arbitrarily, as long as they do not violate the spirit of the present invention, they should also be regarded as the content disclosed in the present invention.

Claims

1. A method for performing similarity routing on a target layout based on the original layout routing information, characterized in that, Includes the following steps: S101. Obtain the routing information of the original layout, wherein the routing information includes the position, physical size, layer and network information of the lines and vias; S102. Obtain the layer information of the target layout and establish a corresponding relationship with the layer information of the original layout, and read the routing constraint information on each layer of the target layout; S103. Based on the routing information of the original layout, extract the topology vector and geometric constraint vector of the original routing; S104. Call the similarity routing algorithm to explore routing. During the exploration process, use the topology vector and geometric constraint vector to optimize the exploration direction. If the routing is successful, proceed to step S107; otherwise, proceed to step S105. S105. Based on the topology vector and the geometric constraint vector, initialize the similar wiring diagram; S106. Iteratively execute the A* search algorithm on the similar wiring graph to search for wiring paths until the wiring solution with the highest similarity is obtained. S107. Draw the wiring diagram on the target layout.

2. The method for performing similarity routing on a target layout based on original layout routing information according to claim 1, characterized in that, In step S102, the wiring constraint information includes minimum line width and minimum line spacing.

3. The method for performing similarity routing on a target layout based on original layout routing information according to claim 1, characterized in that, Step S105 includes: S1051. Using the center of the topological structure vector and the geometric constraint vector as a reference, the similarity gradient of the mesh in the similar wiring diagram is made to spread outward in a Gaussian distribution. S1052. Based on the similarity gradient, assign a higher candidate priority to the grid where the center is located, and complete the initialization of the similar wiring diagram.

4. The method for performing similarity routing on a target layout based on original layout routing information according to claim 3, characterized in that, Step S106 includes: S1061. In the heuristic function of the A* search algorithm, a similarity evaluation term calculated based on the topological structure vector and geometric constraint vector is introduced to guide the search path to preferentially approach the center of the topological structure vector and geometric constraint vector and automatically bypass obstacles.

5. The method for performing similarity routing on a target layout based on original layout routing information according to claim 1, characterized in that, In step S104, the line exploration similarity routing algorithm is a meshless routing algorithm. In the process of selecting escape points, the meshless routing algorithm combines the topological structure vector and the geometric constraint vector to control the direction of the escape points.

6. The method for performing similarity routing on a target layout based on original layout routing information according to claim 4, characterized in that, In step S106, the similarity of the wiring results is evaluated using a similarity measurement formula, which is: ； ； ； in, For topological similarity, For geometric similarity, Indicates wiring reference, Indicates similar routing results. Indicates the final similarity. Indicates the intermediate similarity. The weights for topological similarity. The weights for geometric similarity.

7. A method for performing similarity routing on a target layout based on original layout routing information according to claim 6, characterized in that, The topological similarity The calculation formula is: ； in, Topology sequence for wiring reference Topology sequence of wiring results The Levinstein distance between them For the length of the wiring reference topology sequence, The length of the topology sequence of the wiring result.

8. A method for performing similarity routing on a target layout based on original layout routing information according to claim 6, characterized in that, The geometric similarity The calculation formula is: ； ； in, For length similarity, For the similarity of the included angle, The weights for length similarity. The weights for the similarity of the included angles.

9. A method for performing similarity routing on a target layout based on original layout routing information according to claim 8, characterized in that, The length similarity The calculation formula is: ； in, A vector representing the proportion of line segment lengths for routing reference. This is the vector representing the length ratio of the line segments in the wiring result.

10. A method for performing similarity routing on a target layout based on original layout routing information according to claim 8, characterized in that, The similarity of the included angle The calculation formula is: ； ； in, The vector representing the angle between any two line segments in the routing reference. This is the vector of the angle between any two line segments in the wiring result. Let θ be the minimum length among θ and ϕ.

Images