 |
|
|
Current Projects | Completed Projects | |
|
Research Current Projects
Incorporating Statistical Variability into Circuit-Electromagnetic Modeling of RF and Analog Systems Students: Arun Sathanur and Ritochit Chakraborty Statistical analysis of RF circuits poses new challenges in combining the effects of process variations in both active and passive devices. Active devices include MOS transistors that are susceptible to a wide range of process parameters such as gate oxide thickness, zero body-bias threshold voltage and reduction in channel length to name a few. Correlated process variables exacerbate the problem. We investigate the effects of process variability on performance metrics of RF circuits such as gain, input/output return loss and reverse isolation. By furnishing PDFs of circuit performance measures, our studies highlight design centering approaches for RF circuits. Key Papers and Presentations A.V. Sathanur, R. Chakraborty, and V. Jandhyala, “A hierarchical electromagnetic-circuit technique for statistical analysis of RF circuits in the spectral domain,” IEEE Transactions on Microwave Theory and Techniques, vol. 57, no.4, pp. 796-806, April 2009. [pdf] A.V. Sathanur, R. Chakraborty, V. Jandhyala, F. Ling, D. Gope, and S. Chakraborty, “An accurate hierarchical electromagnetic-circuit technique for statistical analysis of RF circuits,” invited for special session on on-chip electromagnetics, Proceedings of the IEEE Antennas and Propagation International Symposium, pp, 1-4, San Diego, July 2008. [pdf] A.V. Sathanur, R. Chakraborty , and V. Jandhyala, "Accurate statistical analysis of a differential low noise amplifier using a combined SPICE-field solver approach," IEEE ISCAS International Symposium on Circuits and Systems, pp. 884-887, Seattle , May 2008. [pdf] A.V. Sathanur, R. Chakraborty, and V. Jandhyala, “Statistical analysis of RF circuits using combined circuit simulator – full wave field solver approach,” Proceedings of the IEEE/ACM ICCAD International Conference on Computer-Aided Design, pp. 11-17, San Jose, CA, November 2007. [pdf] A.V. Sathanur and V. Jandhyala, “Statistical study of on-chip spiral inductors using integral equation based full-wave electromagnetic analysis,” Proceedings of the IEEE Antennas and Propagation International Symposium, pp. 5757-5760, Hawaii, June 2007. [pdf]
|
Statistical variability pervades every aspect of microelectronics fabrication and design. As design gaps increase, there is a growing need to design optimally in the presence of statistical variability. Eventually, device, sub-system, and system yield prediction also rely on appropriate modeling of the effects of statistical variability. From a simulation viewpoint, cross-layer circuit-electromagnetic modeling is challenging even without parametric and variability modeling. In this work, we aim to build physics-based approaches to handling cross-layer variability at the EM and circuit levels. The nature of statistical variability across multiple physics layers, and its representation at the appropriate physical layer forms part of the challenge. We create statistical models at the physical layer nearest to the physical causation for the variability, therefore avoiding fallacies such as assumed Gaussianity for obviously non-Gaussian entities. Our proposed approach is aimed at helping designers optimize designs robustly in the presence of system variability, and also to predict device and system yields.