next up previous contents
Next: 3.3.0.1 Design of Experiments Up: 3. Statistical Analysis Previous: 3.2.6.4 F Test

3.3 Framework Integration

In TCAD applications these statistical methods are applied to simulation results. In contrast to the analysis of observed experiments these have some major advantages.

Simulated results can be repeated with high accuracy and environmental influences can be excluded. But on the other side the analysis is done with a numerical model of the real phenomenon and the user has to make sure that this model adequately represents the real effects.

Especially complex simulations, like the modeling of the manufacturing stages of the semiconductor devices, require many stages which are usually executed by a simulation environment. They are capable to set input parameters and to extract response from the resulting data, see Figure 3.7. The framework modules, responsible for the program execution, like the sequencer and the qman pass the values of the controls to the simulator and make the simulation results accessible for other modules.

Figure 3.7: Integration of the statistical analysis tools into the framework.
\includegraphics[width=11.cm]{graphics/analysis.eps}

The Design of Experiments and Response Surface Methodology modules are implemented in two separate programs. So these modules can be integrated into a framework and used for stand alone applications.




next up previous contents
Next: 3.3.0.1 Design of Experiments Up: 3. Statistical Analysis Previous: 3.2.6.4 F Test

R. Plasun