摘要
本文在布线的相交图模型基础上,利用离散型的Hopfield网络解决了相交图的最大切割问题,从而解决了双层布线的分层及通孔最少化问题。新算法考虑了许多来自实际问题的约束,并对大量的布线实例进行了验证。
Based on the intersection graph model of routing, we solved the maximum cut problem ofintersection graphs, thus the layer assignment and the via minimization problem for two layersrouting by means of discrete Hopfield neural network. The new algorithm took manypractical constraints into consideration and was verified by a number of routing examples.
关键词
神经网络
集成电路
多层布线
通孔
Artificial intelligence
Manufacture
Neural networks
Optimization
