As MOSFETs are scaled to the sub-micron regime, the doping profile becomes increasingly important affecting the device performance. Profile optimization can be performed either on the process level or on the profile-abstraction level, which means that instead of varying certain process parameters, analytical models are used to describe the doping profile and the model parameters are optimized. This method provides higher speed because the process simulation steps are replaced by a fast analytical device generator. The demand from semiconductor industry for such methods has increased rapidly in the past few years offering the possibility of fitting simulated doping profiles to measured device data (inverse modeling [10,28]).
In this section the doping profile of a specified device structure is optimized to achieve certain performance improvements. Analytical doping profiles are generated by a template based device generator. Parameters are global technology values like the geometry and doping level of a layer. These parameters are used as control parameters for the optimization process.