In-Field Recovery of RF Circuits from Wearout Based Performance Degradation

Doohwang Chang, Jennifer N. Kitchen, Sayfe Kiaei, Sule Ozev

Research output: Contribution to journalArticlepeer-review

7 Scopus citations


Performance failure due to aging is an increasing concern for RF circuits. While most aging studies are focused on the concept of mean-time-to-failure, for analog circuits, aging results in continuous degradation in performance before it causes catastrophic failures. In this paper, we present a methodology for monitoring and recovering the performance of RF circuits in the field at little or no performance penalty. The proposed technique is based on two phases: During the design time, degradation profiles of the aged circuit are obtained through simulations. From these profiles, we identify reliability hotspots and focus on monitoring these components, and recovering from the effects of their aging. After deployment, an on-chip monitor circuit is periodically activated and its results are used to trigger the recovery mechanism if necessary. The recovery mechanism is designed to offset the degradation in the reliability hotspots to enhance the lifetime of the circuit. Lifetime is defined as the point where at least one specification of the circuit fails due to aging degradation. A Low noise amplifier (LNA) is fabricated as a case study to demonstrate that the lifetime can be enhanced by the proposed monitoring and recovery techniques.

Original languageEnglish (US)
Article number8003495
Pages (from-to)442-452
Number of pages11
JournalIEEE Transactions on Emerging Topics in Computing
Issue number2
StatePublished - Apr 1 2020


  • Hot-carrier injection (HCI)
  • RF reliability
  • in-field monitoring and recovery
  • lifetime enhancement
  • low-noise amplifier (LNA)

ASJC Scopus subject areas

  • Computer Science (miscellaneous)
  • Information Systems
  • Human-Computer Interaction
  • Computer Science Applications


Dive into the research topics of 'In-Field Recovery of RF Circuits from Wearout Based Performance Degradation'. Together they form a unique fingerprint.

Cite this