View in EDS

A machine learning-based design automation framework for differential mmWave LNAs.

Saved in:
Bibliographic Details
Title: A machine learning-based design automation framework for differential mmWave LNAs.
Authors: Michailidis, Anastasios1 (AUTHOR) anamicha@physics.auth.gr, Sad, Christos (AUTHOR), Noulis, Thomas (AUTHOR), Siozios, Kostas (AUTHOR)
Source: Integration: The VLSI Journal. Sep2025, Vol. 104, pN.PAG-N.PAG. 1p.
Subjects: Low noise amplifiers, Impedance matching, Differential amplifiers, Complementary metal oxide semiconductors, Passive components
Abstract: In this work, a design methodology of a single-stage differential narrow-band mmWave LNA is presented using a novel full-design automation framework. A differential LNA test case vehicle was designed using a 22 nm FDSOI CMOS process and the ML framework was developed according to this specific process. The proposed framework is based on circuit optimization loops regarding noise figure, gain and impedance matching operating frequency. The proposed framework is capable of generating differential LNA designs with ≥ 99 % input/output matching efficiency, low noise < 4. 4 dB, high gain > 14 dB, high linearity > − 19 dBm, for frequencies of 32-91 GHz. • Fast and automated ML-based characterization of selected process limits. • Automated ML-based active component optimization for minimum noise and enhanced gain. • Automated ML-based passive component optimization and on-chip balun geometry synthesis for high input/output impedance matching efficiency. • High-performance differential LNA synthesis for a wide spectrum of mmWave applications. • A fully-automated ML framework for mmWave LNA synthesis, reducing the design cycle from months to few hours. [ABSTRACT FROM AUTHOR]
Copyright of Integration: The VLSI Journal is the property of Elsevier B.V. and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database: Engineering Source
Description
Abstract:In this work, a design methodology of a single-stage differential narrow-band mmWave LNA is presented using a novel full-design automation framework. A differential LNA test case vehicle was designed using a 22 nm FDSOI CMOS process and the ML framework was developed according to this specific process. The proposed framework is based on circuit optimization loops regarding noise figure, gain and impedance matching operating frequency. The proposed framework is capable of generating differential LNA designs with ≥ 99 % input/output matching efficiency, low noise < 4. 4 dB, high gain > 14 dB, high linearity > − 19 dBm, for frequencies of 32-91 GHz. • Fast and automated ML-based characterization of selected process limits. • Automated ML-based active component optimization for minimum noise and enhanced gain. • Automated ML-based passive component optimization and on-chip balun geometry synthesis for high input/output impedance matching efficiency. • High-performance differential LNA synthesis for a wide spectrum of mmWave applications. • A fully-automated ML framework for mmWave LNA synthesis, reducing the design cycle from months to few hours. [ABSTRACT FROM AUTHOR]
ISSN:01679260
DOI:10.1016/j.vlsi.2025.102435