ANN-Based Spiral Inductor Parameter Extraction and Layout Re-Design

A neural network approach is presented for modeling and characterization of on-chip copper spiral inductors. The approach involves the creation of neural network models to map 3D multi-level spiral inductor geometric characteristics to SPICE equivalent circuit parameters. The neural network replaces computationally expensive FEM-based extraction and field solution. The approach is especially attractive because it is capable of accurately and efficiently predicting important inductor characteristics such as self-inductance, Q-factor, selfresonant frequency and parasitic resistance and capacitance. It also offers substantial computational savings over field solution evaluation of neural model required on average less than 5% of the cpu time required for field solution.