An Aging Small-Signal Model for Degradation Prediction of Microwave Heterojunction Bipolar Transistor <i>S</i>-Parameters Based on Prior Knowledge Neural Network
Lin Cheng,
Hongliang Lu,
Silu Yan,
Chen Liu,
Jiantao Qiao,
Junjun Qi,
Wei Cheng,
Yimen Zhang,
Yuming Zhang
Affiliations
Lin Cheng
Key Laboratory for Wide Band Gap Semiconductor Materials and Devices of Education Ministry, School of Microelectronics, Xidian University, Xi’an 710071, China
Hongliang Lu
Key Laboratory for Wide Band Gap Semiconductor Materials and Devices of Education Ministry, School of Microelectronics, Xidian University, Xi’an 710071, China
Silu Yan
Key Laboratory for Wide Band Gap Semiconductor Materials and Devices of Education Ministry, School of Microelectronics, Xidian University, Xi’an 710071, China
Chen Liu
Key Laboratory for Wide Band Gap Semiconductor Materials and Devices of Education Ministry, School of Microelectronics, Xidian University, Xi’an 710071, China
Jiantao Qiao
Key Laboratory for Wide Band Gap Semiconductor Materials and Devices of Education Ministry, School of Microelectronics, Xidian University, Xi’an 710071, China
Junjun Qi
Key Laboratory for Wide Band Gap Semiconductor Materials and Devices of Education Ministry, School of Microelectronics, Xidian University, Xi’an 710071, China
Wei Cheng
Science and Technology on Monolithic Integrated Circuits and Modules Laboratory, Nanjing Electronic Devices Institute, Nanjing 210016, China
Yimen Zhang
Key Laboratory for Wide Band Gap Semiconductor Materials and Devices of Education Ministry, School of Microelectronics, Xidian University, Xi’an 710071, China
Yuming Zhang
Key Laboratory for Wide Band Gap Semiconductor Materials and Devices of Education Ministry, School of Microelectronics, Xidian University, Xi’an 710071, China
In this paper, an aging small-signal model for degradation prediction of microwave heterojunction bipolar transistor (HBT) S-parameters based on prior knowledge neural networks (PKNNs) is explored. A dual-extreme learning machine (D-ELM) structure with an adaptive genetic algorithm (AGA) optimization process is used to simulate the fresh S-parameters of InP HBT devices and the degradation of S-parameters after accelerated aging, respectively. In addition to the reliability parametric inputs of the original aging problem, the S-parameter degradation trend obtained from the aging small-signal equivalent circuit is used as additional information to inject into the D-ELM structure. Good agreement was achieved between measured and predicted results of the degradation of S-parameters within a frequency range of 0.1 to 40 GHz.
