how about soliton....i think its one of solution

Hi Akhmad,

 Thank you for the suggestion. Soliton is definitely one solution but i want to use it at the final implementation. i want to know some parameters for example frequency separation of the two tones for a specific bit rate of the system.

Regards,

Dhiman

Dhiman: Can you explain in more detail what you are trying to do?  What power levels, propagation distances and signal wavelengths?

Are you concerned about distortion of the composite dual wavelength waveform, or will you separate the two wavelengths and consider the two signals independently?  Is this a single span system or are there multiple optically amplified spans?

At 10 Gbit/s there is significant single-channel distortion due to chromatic dispersion after 1000 ps/nm fibre dispersion.  This corresponds to around 60 km of standard fibre at 1550 nm, but several hundred km near the 1300 nm dispersion zero wavelength.  The impact of dispersion varies quadratically with modulation rate, so dispersion tolerance is 16x lower at 40 Gbit/s.

How sensitive to phase noise are you?  Are you using direct detection of the signal envelope, coherent detection, or some other scheme with increased sensitivity to phase noise?

One source of distortion is cross phase modulation, XPM.  The fibre non-linear coefficient of standard fibre at 1550 nm is around 0.0012 /W/m,  so for 10 mW signal modulation the resultant phase modulation over 20 km in a co-propagating signal at a closely spaced wavelength would be 0.24 to 0.48 radian, depending on the relative polarisations of the signals.  At high modulation rates and for increasing wavelength separation, the induced phase modulation depth is reduced by walk-off arising from the difference in propagation velocity.  In the presence of chromatic dispersion, there will also be phase to amplitude conversion. The impact will depend strongly on the modulation bandwidth, fibre dispersion and channel separation.

Is intra-channel waveform distortion due to Kerr non-linearity a concern?  Four wave mixing could contribute at high powers, but (depending on the dispersion regime) XPM is likely to dominate at moderate powers in a 2-wavelength system.

Soliton propagation relies on the balance between chromatic dispersion and Kerr non-linearity, and for best results requires tight control of the fibre dispersion and signal power.  With closely spaced WDM channels XPM-induced timing jitter can be a problem.

Are you still targeting a channel separation of 40 GHz?  It is difficult to advise without more information on how the signals are modulated and detected, and what you are trying to measure.

Are you able to apply dispersion compensation at the transmitter or receiver?  This can mitigate some of the effects of fibre dispersion and Kerr effect non-linearity.

Hello sir, 

I want to use nearby 0 dBm (i.e. 1 mw) laser power, 

propagation distance around 50-60 km. 

wavelength = 1552 nm

want to transmit the dual wavelength signal at the same time through the single channel so that it will be suitable use of one wavelength for uplink and other for downlink.

I am using Direct-detection scheme and DQPSK modulation format.

I need to study more about this cross phase modulation and Kerr-nonlinearity. I need some more time to tackle this issue.

in terms of channel separation presently the two mm-wave that I am using for modulating the optical carrier are of 50GHz and 60 GHz  i.e. channel separation is of 10 GHz.

Presently I am not using any kind of dispersion compensation technique. 

Thank you so much for providing me with valuable information, and helpful to me while modifying the design.

Regards,

Dhiman

Dhiman, just so I understand what you are doing:

You are not modulating the optical carrier directly.  You amplitude modulate an optical carrier with a microwave signal.  The microwave carrier is centred on either 50 GHz or 60 GHz, and carries DQPSK modulation.  The modulation envelope is recovered by simple direct detection at the receiver, and data extracted by RF processing of the microwave sub-carrier.

Alternatively, do you modulate a single 1552 nm optical carrier with both 50 GHz and 60 GHz microwave sub-carriers?  The will produce a more complex wide bandwidth composite waveform. 

In either case a 50-60 GHz modulation envelope will be severely degraded by linear chromatic dispersion after propagating through a few kilometres of standard single mode fibre.

Do the up-link and down link signals propagate in opposite directions?  If this is the case, then cross-phase modulation will not be important.

 What is the bit rate or symbol rate for the data carried by the microwave sub-carriers?  Why not modulate independent 1551.72 nm (193200 GHz) and 1552.12 nm (193150 GHz) sources using IMDD or DQPSK modulation?  These are standard ITU wavelengths, and can be separated using of-the-shelf 50 GHz DWDM filters.  For bi-directional propagation, you could use a directional coupler or circulator.

At 0 dBm launch power XPM crosstalk will introduce some phase noise, but might not be a serious problem.

Regards,

Alan.

Sir,

both the wavelength will travel in one direction, then one wavelength is extracted and used to carry the uplink data. So cross phase modulation must be considered.

presently using 193.1 THz,

yes for sure I have tried circulator and good results also.

Thanking you..

Regards,

Dhiman

Dhiman,

I still don't fully understand the spectrum of the signal you are trying to propagate, and how it is modified along the link.

If you amplitude modulate the signal at 50 GHz and 60 GHz, then you will probably get tones at 193.04, 193.05, 193.10, 193.15, 193.16 THz - a spectral width of 120 GHz - with more tones possible depending on the modulation scheme. How is the modulation applied?  Do you apply CW modulation and adjust the amplitude to suppress the 193.1 THz carrier, or is this always present?  Do the +/-50 GHz tones propagate over part of the route with additional tones added when the uplink modulation is applied?

Do the 50 GHz and 60 GHz microwave signals carry DQPSK data, or do you extract CW tones and use a pair of Mach-Zehnder modulators to apply the modulation to a single spectral line? 

What is the symbol rate for the data modulation?

As I said previously, if you are trying to propagate a 60 GHz microwave sub-carrier, it will be heavily distorted within a few km. 

If you modulate an optical carrier directly with DQPSK at 10 GBd or less, you should be able to propagate over 60 km SMF with tolerable distortion, provided you optically filter unwanted spectral components at the receiver.

Regards,

Alan.