Abstract
Process-induced variability is a growing concern in the design of analog circuits, and in particular for monolithic microwave integrated circuits (MMICs) targeting the 5G and 6G communication systems. The RF and microwave (MW) technologies developed for the deployment of these communication systems exploit devices whose dimension is now well below 100 nm, featuring an increasing variability due to the fabrication process tolerances and the inherent statistical behavior of matter at the nanoscale. In this scenario, variability analysis must be incorporated into circuit design and optimization, with ad hoc models retaining a direct link to the fabrication process and addressing typical MMIC nonlinear applications like power amplification and frequency mixing. This paper presents a flexible procedure to extract black-box models from accurate physics-based simulations, namely TCAD analysis of the active devices and EM simulations for the passive structures, incorporating the dependence on the most relevant fabrication process parameters. We discuss several approaches to extract these models and compare them to highlight their features, both in terms of accuracy and of ease of extraction. We detail how these models can be implemented into EDA tools typically used for RF and MMIC design, allowing for fast and accurate statistical and yield analysis. We demonstrate the proposed approaches extracting the black-box models for the building blocks of a power amplifier in a GaAs technology for X-band applications.
Highlights
The foreseen transition to 6G communication systems calls for increased operation frequency and bandwidth along with reduced power dissipation and high efficiency, opening the way to the exploitation of new technologies and devices
A distinctive feature of microwave circuit design is the need for accurate modeling of both active devices and passive structures, which are in many cases implemented in semi-lumped form, i.e., adopting both distributed and lumped (MIM capacitors and spiral inductors) elements
We show that black-box models are well suited to bridge this gap, as they directly translate the physical simulations into electronic design automation (EDA) circuit design environments
Summary
The foreseen transition to 6G communication systems (and beyond) calls for increased operation frequency and bandwidth along with reduced power dissipation and high efficiency, opening the way to the exploitation of new technologies and devices. The random nature of the technological variations, either linked to the granularity of matter at the nanometer scale or to the fabrication process tolerances, makes statistical analysis a fundamental tool for the design and optimization of a microwave stage In this perspective, the designer must be aware that circuit optimization relying on the nominal device parameters only may result in being blurred out, or even utterly impaired, when the technological spread is taken into account. As for passive structures, physics-based simulations through calibrated technology CAD (TCAD) would represent the ideal framework to incorporate PIV into microwave design, but EDA tools do not allow for co-simulation of the active device physical model into the circuit-level design flow, mainly due to the numerical burden of the nonlinear physical model (e.g., the drift-diffusion model) solution. We focus in particular on models allowing for a flexible implementation into the most common EDA tools for RF and MMIC design, and a fast, yet accurate, statistical circuit analysis of PIV
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
