X-Architecture Steiner Tree Construction via Particle Swarm Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing X-architecture Steiner minimum tree construction algorithms face high time complexity and are prone to local extremums, failing to optimize wirelength effectively in very large scale integration routing, especially as problem scales increase.
Innovation Solution
A two-step method using particle swarm optimization with social learning discrete particle swarm search and local topological optimization to construct an X-architecture Steiner tree with minimum wirelength, incorporating edge-vertex coding, chaotic mutation, and social learning strategies to reduce redundancies and improve routing quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If accurate algorithms are used for X-architecture Steiner minimum tree construction, then routing quality is improved, but time complexity increases drastically
Solution Approach 1:
The patent segments the Steiner minimum tree construction into two distinct phases: (1) topology construction phase using heuristic algorithms to build an initial routing tree, and (2) optimization phase using particle swarm optimization to refine the topology. This segmentation allows the algorithm to balance between computational efficiency and routing quality by handling different aspects of the problem separately.
Solution Approach 2:
The patent applies local quality by focusing optimization efforts on critical parts of the routing tree. The particle swarm optimization specifically targets the topology structure and Steiner point positions, while keeping other aspects fixed during each iteration. This localized optimization approach improves routing quality without requiring complete re-optimization of the entire tree structure.
2Loss of time
If heuristic algorithms are used for X-architecture Steiner minimum tree construction, then time complexity is reduced, but the algorithm is prone to local extremums
Solution Approach 1:
The patent performs preliminary action by first constructing a complete initial routing tree topology using heuristic algorithms before applying particle swarm optimization. This preliminary construction ensures that a feasible solution exists and provides a good starting point for the subsequent optimization phase, preventing the algorithm from getting stuck in poor local extremums.
Solution Approach 2:
The patent implements feedback mechanisms through the particle swarm optimization process, where each particle's position represents a potential routing topology and its fitness is evaluated based on wirelength. The algorithm continuously updates particle positions based on feedback from fitness evaluations and social learning from other particles, enabling escape from local extremums through iterative improvement.
3Ease of manufacture
If Manhattan architecture is used for routing, then routing directions are constrained to horizontal and vertical, but optimization of interconnection wire length becomes more difficult
Solution Approach 1:
The patent applies asymmetry by breaking the conventional Manhattan architecture's symmetric horizontal-vertical constraint and introducing asymmetric 45-degree diagonal routing directions. This creates an X-architecture where routing can occur in multiple directions (horizontal, vertical, and two diagonal directions), allowing more flexible wirelength optimization while maintaining manufacturability through standardized routing layers.
Data Source
AI summary
The invention relates to the technical field of computer-aided design of integrated circuits, and provides a two-step X-architecture Steiner minimum tree construction method for very large scale integration (VLSI). Based on the advantages of an X-architecture model and a particle swarm optimization technique, the method is implemented through two steps: (1) the stage of social learning discrete particle swarm search, which comprises: using an edge-vertex encoding strategy capable of maintaining optimal topological information of particles, designing a fitness function taking wirelength into consideration; and using a chaotic decreasing mutation strategy and a new social learning strategy to design a new discrete particle swarm update formula; and (2) a stage of wirelength optimization, which comprises: designing a local topological optimization strategy to minimize the wirelength of an X-architecture Steiner tree. The method guarantees short total wirelength of nets and has high stability, thus being able to construct a high-quality X-architecture Steiner minimum tree.