The necessary information to perform the WAN design optimization are organized into seven databases:

• Applications;
• Locations of the sources and destination of traffic;
• Possible concentration nodes;
• Servers locations;
• Interconnection costs;
• Hardware modularity;
• Possible interconnections

These files are the ones, that we call primary files. They are populated by us directly or by our support software. All other files are generated based on these ones.

We have two families of what we call support software:

• "Curve readers" - Software that help us to format the information regards applications traffic based on raw data originated from measurements. (telephone bills and circuits measurements graphics);
• "Non linear regression" - Through non-linear regression techniques we are able to identify rules for prices' definition.