Investigation and development of a novel metrology standard for the measurement of relative intensity noise and frequency chirp of DFB lasers in optical networks

Cox, Michael Christopher (2007) Investigation and development of a novel metrology standard for the measurement of relative intensity noise and frequency chirp of DFB lasers in optical networks. Doctoral thesis, London Metropolitan University.


Laser diode Relative Intensity Noise (RIN) metrology capabilities have been developed and demonstrated, providing significantly improved sensitivity and accuracy than existing methods. This is a key parameter for laser components used in telecommunication systems due to the drive to reduce inherent noise, improve overall signal-to-noise ratio and thus increasing achievable communication link length and reduce required component specification.

The novel use of the demonstrated reference noise source has shown significant advantages, achieving improved sensitivity, increase measurement accuracy as low as ±ldB and simplifying the system calibration methodology thus improving flexibility. Laser RIN of between 10 to 14dB below the Shot RIN have been shown (typically -170dB/Hz) which is a direct result of the improved system sensitivity.

The developed system is based on an Neodinium Yag (Nd:Yg) 1319nm ring laser, essentially providing a ‘cold’ reference source, in a similar manner to that used in rf electrical metrology. High finesse Nd:Yag lasers inherently emit significantly small amounts of AM/FM noise being spectrally very pure. This is enforced further by the nonplanar ring oscillator (NPRO) technology providing superior line width stability. The selection of a high output power version allows optical attenuation to be applied which provides a further 20dB attenuation of system spontaneous noise. Application of the ‘flat’ low noise optical rf noise source from 10MHz to 20GHz has been demonstrated for the first time in optical rf metrology, providing a calculable reference traceable via the incident optical power received. Due to the simplistic nature of this approach, system calibration can be measured for each RIN measurement performed, reducing measurement uncertainty associated with rf miss-match, system linearity and loss.

High specification components have been assessed individually and in the combined system indicating overall system noise figure of 2 to 3dB over the 10MHz to 20GHz frequency range (-171dBm to -172dBm), some 4 to 5dB better than previously reported. Good agreement has been shown comparing thermal noise subtraction and 3dB rise techniques confirming the theoretical dominance of the rf amplifiers noise figure.

Application of narrow band optical signals, by way of directly modulating a DFB laser and using an external optical modulator has confirmed the noise sensitivities found previously and also the theoretical 2dB wideband offset reported in literature.

Measurement methodology has been reported, describing the three stage approach and simple derivation of RIN. Extensive assessment of measurement uncertainty contributions are reported complying with UKAS standard guidelines for random and systematic terms. The overall uncertainty budget drawn from this is dynamically linked to the level of RIN under test. Application of the reference laser has additional benefits in allowing linearity contributions to be realised.

Intercomparisons with a commercially available RIN measurement system highlighted the sensitivity advantages of the reference technique. A number of short comings of the traditional approach in terms of systematic errors, such as the spontaneous loss over attenuation are reported.

Frequency excursion (chirp) metrology capabilities have been explored, developed and demonstrated. The final system is based on the gated delayed self homodyning interferometer. Inclusion of Faraday rotating mirrors has been reported showing favourable results, reducing systematic uncertainty by 7% at low modulation frequencies. These mirrors, one on each arm of the interferometer, effectively cancel out any fibre birefringence within the interferometer and also negate the need for a polariser to maximise the output signal.

Initial accuracies of ±10% have been achieved traceable to national standards. Intercomparisons with a commercially available interferometer show a 7 to lOdB more amplitude sensitivity thus improving the resolution capability, aiding line width and chirp determination. Analysis has been limited to low modulation frequencies due to equipment delivery limitations. The reference source has been utilised to provide modulation level calibration.

Concept for high modulation frequency testing has been reported based on tracking the relative ratio of two available Bessel peaks. This has been modelled via classic FM theory, knowing the line width, and modulation index to obtain a best fit. This technique has advantage in being able to operate without the need to meet a ‘null’.

851207.pdf - Published Version

Download (94MB) | Preview


Downloads per month over past year

Downloads each year

View Item View Item