Highly–sensitive phase and frequency noise measurement technique using Bayesian filtering

Darko Zibar*, Hou-Man Chin, Yeyu Tong, Nitin Jain, Joel Guo, Lin Chang, Tobias Gehring, John E. Bowers, Ulrik Lund Andersen

*Corresponding author for this work

Research output: Contribution to journalJournal articleResearchpeer-review

530 Downloads (Pure)


Spectral purity of laser sources is typically investigated using phase or frequency noise measurements, which require extraction of the optical phase. This is a challenging task if the signal–to–noise–ratio (SNR) of the spectral line or the linewidth–to–noise–ratio (LNR) are not sufficiently high. In this paper, we present a statistically optimal method for optical phase noise measurement that relies on coherent detection and Bayesian filtering. The proposed method offers a record sensitivity, as the optical phase is measured at a signal power of -75 dBm (SNR of -11 dB in 1.1 GHz receiver bandwidth). Practically, this means that the phase noise measurements are, up to a high–degree, not limited by the measurement noise floor. This allows measurements down to -200 dB rad2/Hz and up to 10 GHz, which is useful when measuring the Schawlow–Townes (quantum noise limited) laser linewidth. Finally, the estimated optical phase is highly accurate allowing for quantum limited signal demodulation. The method thus holds the potential to become a reference measurement tool.
Original languageEnglish
JournalI E E E Photonics Technology Letters
Issue number23
Pages (from-to)1866 - 1869
Publication statusPublished - 2019


  • Phase noise
  • Lasers
  • Frequency combs
  • Bayesian filtering
  • Machine learning


Dive into the research topics of 'Highly–sensitive phase and frequency noise measurement technique using Bayesian filtering'. Together they form a unique fingerprint.

Cite this