Zwei neue Strategien zur Fuzzy-Bipartitionierung von Schaltungs-Hypergraphen

The paper presents two new strategies for bipartitioning circuits based on fuzzy set theory. For this a new fuzzy net-cut model has been developed treating hypergraphs without splitting multi-pin-nets into two-pin-nets. The first approach is based on a stochastic neural network consisting of continuous Hopfield neurons. The second algorithm is derived from statistical physics modelling the circuit graph as a highly coupled spin system.