Title :
Graphical model validation methods for analog and mixedsignal electronic circuits design
Author :
Hamad, Husam ; Al-Smadi, Adnan ; Ijjeh, Abdallah
Author_Institution :
Biomed. Eng. Dept., Jordan Univ. of Sci. & Technol., Jordan
Abstract :
Simplified models are used to replace SPICE-based simulation models in analog and mixed-signal electronic circuits design for increased computational efficiency. The increase in computational efficiency usually comes at the expense of reduced model accuracies by comparison to simulation models. Graphical model validation methods can be used to display model and simulation data to complement validation statistics such as root mean square (RMS) error. In this paper, we present three plot types to display data to validate models in analog integrated circuits. These are the so-called circle plots, acceptance-score-distribution (ASD) plots, and ordinal plots. These plots are general and easy to generate using the same validation sample used to calculate other statistics, e.g., RMS errors; no extra simulation expenses are incurred. Furthermore, unlike quantitative statistics used for model validation, the methods presented require no statistical assumptions for the data that can be displayed.
Keywords :
electronic engineering computing; integrated circuit modelling; mean square error methods; mixed analogue-digital integrated circuits; acceptance-score-distribution plot; analog circuit design; analog integrated circuits; circle plot; graphical model validation; mixed-signal electronic circuit design; ordinal plot; root mean square error; Analog integrated circuits; Circuit simulation; Computational efficiency; Computational modeling; Displays; Electronic circuits; Error analysis; Graphical models; Integrated circuit modeling; Root mean square; Analog circuits modeling; model validation;
Conference_Titel :
Microelectronics, 2008. ICM 2008. International Conference on
Conference_Location :
Sharjah
Print_ISBN :
978-1-4244-2369-9
Electronic_ISBN :
978-1-4244-2370-5
DOI :
10.1109/ICM.2008.5393849