IEEE P1076.1 Working Group - Table-Driven Modeling Project - Telco Meeting Minutes of December 6, 2010

Status: approved on January 10, 2011

Attendees

• Ernst Christen
• Joachim Haase (chair, meeting convener, note taker)
• Arpad Muranyi
• Thuy Tran

Agenda

• Call to order
• Volunteer fro minutes
• Approval of agenda
• Review and approve minutes from the last meeting
• Real-valued 1-dimensional interpolation - status and next steps
• Requirements for complex-valued 1-dimensional interpolation
• Next meeting
• AOB
• Adjourn

Call to order on December 6, 2010 at 8:05am PDT (17:05 CEST)

Joachim welcomes the participants and starts presenting the meeting slides.

Volunteer for minutes

Joachim takes the notes.

Approval for agenda

Agenda unanimously approved.

Minutes from last meeting

Minutes of November 8, 2010 meeting unanimously approved.

Real-valued 1-dimensional interpolation - status and next steps

Joachim presented the results on combinations of interpolation and extrapolation methods. The results of the previous discussions were considered.

Handling of an unproper number data points in the case of quadratic spline interpolation with periodic extrapolation must be yet defined.

To avoid name space problems Ernst will provide a suggestion how to describe interpolation and extrapolation types using an appropriate prefix.

Requirements for complex-valued 1-dimensional interpolation

Consequences of Touchstone format requirements, types COMPLEX and COMPLEX_POLAR of the VHDL MATH_COMPLEX package and AC analysis statements in SPICE were discussed.

Ernst suggested to focus on interpolation of values of type COMPLEX. Other data representations should be transformed into COMPLEX values or COMPLEX vectors resp.

Values for frequency zero must be handled in an appropriate manner. REAL and IMAGINARY parts are interpolated independently by the same interpolation and extrapolation method resp.

Arpad pointed out that characteristics as passivity that are fulfilled by the given data points should/must be also be valid for the interpolation results. Ernst suggested to distinguish between applications where this is necessary and others where it isn't.

In addition to the real-valued interpolation methods rational interpolation should also be provided. Ernst reminds that copyright problems must be avoided for the reference implementation (for instance concerning Stoer/Bulirsch approach).

Next meeting

The next telco is planned on January 10, 2011 at 8am PDT or 17:00 CEST.

AOB

None.

Adjourn

Joachim adjourns the meeting at 17:55 CEST.

Action items

Ernst:

• Proposal for identifiers for interpolation and extrapolation enumaration type.

Joachim:

• Proposal for the interface of a real-valued function for the 1-dimensional case.
• Proposal for the 1-dimensional complex-valued interpolation.