IEEE802.3aq Channel model ad hoc Task 2: Time-varying study

1
IEEE802.3aq Channel model ad hocTask 2
Time-varying study modal noise

• Meeting notes, 15 July 2004
• Jonathan King

2
Attendees

• Jonathan King, Big Bear Networks
• Tricia Hill, Big Bear Networks
• John Jaeger, Big Bear Networks
• Abhijit Phanse, Scintera
• David Cunningham, Agilent
• Piers Dawe, Agilent
• Gary Schaulov, Agilent
• Yu Lung Chau, IKRI
• Matt Treverso
• Andre van Schyndel, Bookham
• Petre Popescu, Quake
• John Dallesasse, Emcore
• John Ewen, JDSU
• Yi Sun, OFS
• Nick Weiner, Phyworks
• Brian Taylor, Oepic
• Paul Kolesar, Systimax
• Kevin Witt, Vitesse
• contributions from

3
Objectives of meeting

• define goals and tasks
• agree time line
• September input required by task 1

4
Goals for task 2

• Study the impact of time varying effects on the
  LRM draft spec
• Provide input to the time varying component of
  the receiver compliance test
• Set a frequency below which EDC is should track
  (TP3 test rate)
• define tracking 'mask' - magnitude of
  perturbation vs frequency of perturbation
• Measure spectrum of time variation effects
• what is the product expected to cope with in real
  life ? what's already out there ?
• consensus that measuring this for statistically
  significant numbers installed links would be an
  extremely difficult task
• Modeling of extent of time variance of links
• e/g. delay vs mode sets for quasi-continuous
  perturbed link for listed mechanisms
• generate impulse responses for quasi-continuously
  perturbed link for selected launches, over
  Cambridge fibre set (tbc) and a 'standard link'
• Study modal noise of the MMF channel and ensure
  modal noise is properly accounted for in the LRM
  draft spec
• Assess modal noise penalty for different laser
  types

5
Time variation mechanisms

• In approx order of significance of magnitude
• connector offset
• static offset 3.5 average, 1.8 standard
  deviation for 50 micron fibre check if we need
  to scale for 62.5 micron fibre. (95 less than
  7um)
• temporal variability tbd
• fibre perturbation
• import fibre shaker for measurements
• mechanical and polarization
• laser wavelength
• can be modeled with mode solver
• temperature
• very slow, believed trackable, not main focus of
  study
• Rate tbd
• Faster than EQ adaption - noise
• modal noise variance can be estimated using
  previous work
• Slower than adaption rate - trackable

6
Tasks

• Mechanisms determine a placeholder value for
  mechanical vibration rate for connectors - John
  Dallesasse
• Experiment relationship of perturbation spectrum
  to modal noise spectrum - Abhijit Phanse will
  check capability for measurements at scintera
• Modelling of delay vs modes for quasi continuous
  perturbed link for listed mechanisms - Yu
  Sun/Pete Hallemeier
• generate quasi continuous impulse responses for
  perturbed link for selected launches (over
  Cambridge fibre set) , 'standard link'
• Modal noise penalty for different laser types -
  David Cunningham
• Effect of link temperature - Petre Popescu
• Real environments - volunteers tbd
• Measurements of spectrum of time variation
  effects
• Input on what is the product expected to cope
  with in real life?
• what's already out there describing in building
  environments

7
Previous work

• Modal Noise Test Methodology group
• 'Measurement of the optical power penalty due to
  modal noise in multimode fiber laser based links'
• Fibre shaker TIA/EIA-455-203
• Sun_1_0704
• Modelled CL and OSL launch sensitivity to
  connector offset 1028010 link

8
Next meeting

• Wednesday morning 9am PCT, 5pm UK