Using wavelengths outside the telecom spectrum


What applications can the unused wavelengths outside of the Telecom spectrum be used for?


Remy de Boer

Stefan Plug

Contents

  • The project
  • The additional wavelength
  • Proof of concept
  • Tests
  • Conclusion

The Project

  • CWDM
  • passive device
  • from 1270nm to 1610nm
  • 20nm wide channels
Basic CWDM setup

The Project (2)

  • BeetleFiberOptics
  • Low-cost
  • Use of extra wavelength
CWDM setup with extra out-of-band wavelength

The additional wavelength

    Band Descriptor Range [nm]
    O-band Original 1260-1360
    E-band Extended 1360-1460
    S-band Short wavelength 1460-1530
    C-band Conventional 1530-1565
    L-band Long wavelength 1565-1625
    U-band Ultra-long wavelength 1625-1675

The U-band has been defined exclusively for possible maintenance purposes.
Ref: ITU-T manual 2009 P. 134

The U-Band

(1625 to 1675)

Attennuation per λ per km
Ref: Computer networks, fourth edition. 2002
Attenuation per λ with bends
Ref: Reference Guide to Fiber Optic Testing, Second Edition, Volume 1. 2011.2

1625nm seems the logical choice

Proof of concept

  • Inteno XG6746
  • Raspberry Pi
  • 1625nm optics

Full network

"plug and play"

Vlan 10

VLAN 10
  • RP-Switch units
  • IPv4 default addressess
  • RP# SNMP polls only its own switch

Vlan 11

VLAN 11
  • RP-groups
  • IPv6 Link-Local addresses
  • Automatic neighbor discovery script

Vlan 12

VLAN 12

Out-of-band management

Proof of concept: Issues

  • Inteno's closed-source
  • Wrong DDM values


Switch Tx Rx OPM measured Rx
Sw1 -4.61 dBm -0.44 dBm -9.8 dBm
Sw2 -4.86 dBm -0.83 dBm -12.4 dBm

Proof of concept: second set-up

Attenuation network
  • Zyxel properly reads DDM values
  • AttF1-F3 = TxF1 - RxF3
  • Doesn't support 100mbit SPF

Tests

  • Attenuation
  • Stability

Attenuation test

OTDR base test, 1550nm, 25 Km, 7.009 dB

Wavelength 1M 25KM
1550nm -0.05 dB 8.3625 dB
1610nm 0.0875 dB 7.975 dB
1625nm 2.275 dB 14.325 dB

Wavelength 1M 25KM
1550nm 0.4425 dB 9.2325 dB
1610nm 0.62375 dB 9.2825 dB
1625nm 2.25125 dB 14.95625 dB

Stability test

10-mbit, 15 hours

Total of 62.9GB transferred

45918348 packets

RP1: iperf -suV

RP2: iperf -c fe80::ba27:eb27:ebff:fe58:69e2%eth0.11 -V -b 10000000 -t 54000

Conclusion

Altough 1625nm is more sensitive to:

  • distance
  • bending
  • temerature changes

Still stable over long distances

FIN

https://github.com/remydb/rp1