hybrid optical amplifier

You’re welcome. The implementation of HOA from papers seems to be as easy as using two optical amplifiers after each other. The important part is its design and effect on the system, and as long as you have a very basic understanding of the software and where to find the optical ampliifers, you can easily implement them. You may find Raman amplifiers harder to set, though. So have a look at optisystem samples regarding Raman amplifiers.

Below I am attaching one Optisystem sample for RA, to which I have added EDFA which increases power by 4 dB. This is NOT from any paper and just shows how the gains are added…just to directly address you question.

Regards

You’re welcome lavanya. I hope it has been helpful.

Hi Swapandeep Kaur,

According to your question about attaching papers, – unfortunately any of us, as an Optiwave forum member, is not able to attached any materials that are not in a public access, i.e. any material that has author’s copyrights. These copyright rules and regulations are expanded to our forum and have been clearly declared by Optiwave Team (Damian if I am not mistaken) in one of the previous forums…

Hi Ravil,

If you are referring to the version of the file that I have attached in this forum page, I have used OptiSystem version 13.0.3 to create the file. If you did not have access to version 7 (as Ajaybeer has demanded), maybe you can kindly attach a screen shot of the design implementation. Unfortunately, I don’t have access to the software right now.

Regards

Hi Ajaybeer,

HOA.osd that I have attached in this topic has been made with OptiSystem version 13, so you would not be able to open it using any older version. I suggest that you download the newest version of the software (which is version 14) from your profile to be able to open the aforementioned OptiSystem file then.

Regards

The average model uses a simplified version of Raman equations to simulate a Raman amplifier (see for example:
B. Min, W. J. Lee, N. Park, “Efficient Formulation of Raman Amplifier Propagation Equations
with Average Power Analysis”, IEEE Photonics Technology Letters, Vol. 12, No. 11, November
2000.).

The dynamic model uses a complete dynamic description of the equation, so it uses more time to compute. It depends on the level of accuracy that you need which model to use – please remember that you can also use the propagation fiber itself for Raman gain simulations!

You are very welcome gaganpreet! I think that for most of the applications the Average Model is ok…can you confirm that simulation time is quicker with it?

Regards,

Alessandro

Ok I understand!

Hi Swadapdeep,

of course hybrid amplification can also be applied to TDM systems…this type of amplification is suitable for any optical signal.
Do you have in mind any particular system?

Best Regards,

Alessandro

Hi Swapandeep,

As Alessandro mentioned, the hybrid optical amplifiers can be used for any system, since their influence is on the optical signal in the fiber, regardless of whatever modulation may have been used. However, I believe cost-efficiency is of importance, for example it may not be economical to use it in a short-distance access system in general.

Regards

Hi, I am not an expert at all on SOA so my previous reply was actually about EDFA-RAMAN.

Do you have any particular design requirements?

For the comparison purposes, you can do the simulations using the amplifiers you want to compare to each other and then compare the results. This is more reliable but takes quite some time. However, you can take a look at the following page which simply introduces and explains about some of the amplifiers you have mentioned, just to have an overall understanding:

http://www.fiber-optic-tutorial.com/comparison-of-different-optical-amplifiers.html

Regards

Hi Gaganpreet,

what you have mentioned is in a general sense about the parametric amplifiers. I believe Swapandeep’s question is more about the use of amplifiers she mentioned for a specific system (which uses TDM) than the general benefits of parametric amplifiers. Therefore, I believe some general ideas about amplifiers whose combinations she is using may be helpful in having a sense about the answer she is going to get doing the simulations.

Regards

I believe the main reason for these amplifiers not being so popular nowadays is the fact that today, the focus is pretty much on other parts of the system rather than the channel to improve performance, so for example a simple EDFA along with complex DSP and coherent detection is preferred in many cases. That’s probably why these amplifiers are not developed that much for today’s systems.

Cheers!

You’re welcome Gaganpreet.

Regards

Allow me to exemplify this: If you are using a 500 km fiber which needs optical amplifier every 100km, you can easily use a 100km fiber followed by an amplifier and put it in a loop using loop control, setting number of loops to 5. Otherwise, you would have to connect 5 pieces of fiber and 5 amplifiers. However, you can certainly simulate the system without using it.

You’re welcome. You do not need to use the amplifiers both before and after the fiber in case you are using loop control. Either you put it before or after the fiber, you can make the comparison. However, you should see why you are using the amplifiers in the system (as power booster, pre-amplifier, or in-line amplifier).

I am sorry Swapandeep I could not find such a component in optiSystem libraries (even though I searched the component in all the libraries). Can you please determine the exact address of the component? Unless you are talking about the EDFA black box component which is available in the same folder.

Please clarify what I mentioned about EDFA in my comment #29619 so that I would be able to explain it. About Raman, I would like to refer you to the good explanation given by Alessandro in this topic in reply #29436 where he states the choice depends on the level of accuracy needed in simulation.

Then I guess we are using different OptiSystem versions! I have EDFA black box component in the OptiSystem library among amplifiers, which does not require detailed design of the amplifier in order to perform with the defined characteristics. Please seeif the ideal EDFA also has the same performance.

I believe you can proceed with comparison by using traveling wave SOA as in-line amplifier as you mentioned in your own comment. Since your work is comparison, you can use EDFA black box (or the EDFA ideal; if it has the same characteristics as those I mentioned for EDFA black box in my former comment).

Regards

You’re welcome Swapandeep. Here I have attached an image of EDFA black box in OptiSystem 13. You can see it and compare it to EDFA ideal in version 7 to see if they perform more or less in the same way if you think it may help you in any way. Maybe they are the same component with the name changed.

Thank you for letting me know of this.

Regards

Nonlinearity factors include self-phase modulation and some other factors that may cause problems for WDM systems. in order to eliminate the effect of this factor, you can click on “nonlinearity” tab in the optical fiber properties to see all the related settings there. Can you explain more about polarization?

If you mean to ask whether polarization can be effected by nonlinear effects or not, it can. In other words, the polarization can depend nonlinearly on the power of the optical field as far as I am concerned, as it can affect the phase of the optical wave propagating through the optical medium in SPM.

the negative effects of nonlinearity can be seen while injecting high amounts of power into the optical fiber. For example, if you do the simulation with and without self-phase modulation with the laser power being 25dB, the negative effects of the optical fiber nonlinearities will be quite obvious.

Provided that the settings for the pump frequency are set correctly, I don’t think any problems should occur using Raman instead of EDFA in your implementation. The pump parameter can be seen as “upper pump reference” parameter in both the dynamic model and the average power model available in library.

You’re welcome Swapandeep. I know about the transmitter you have been using in the .osd file and when I asked you to increase the laser power in my former comments (in order to see the nonlinear effect of SPM), I meant increasing the integrated laser in transmitter. Sorry if it confused you.

Yes, I managed to open your file and I made a comment on how to see the effect of self-phase modulation in your implementation as you had asked. I managed to get the Q-factor of about 4 (with the corresponding BER of about 1e-5). So I can say I don’t see any mistakes in implementing the system.

The reason for the calculations taking so long is the fact that by increasing the input power, the nonlinear effects have kicked in. I suggested increasing power just to see the SPM effect. Please see if you face any problems in getting results by using less input power in your simulations.

Please consider my suggestion in reply #29650.

Regards

Please upload the file with Raman amplifier being used. I will try to help if possible.

Hi Swapandeep,

You have forgotten to pump the Raman amplifier and that is why the problem has happened. In fact, without the pump, Raman amplifier is somehow not considered in the calculations and consequently there is no signal in the output as in your case. Please try this and see if the problem is hopefully over.

Regards

The upper pump limit must be set according to the pump frequencies used. Please go to OptiSystem samples and find the samples in the folder “Raman”. There are two examples in that folder: One regarding average power model and one regarding dynamic model. The average power model will help you set up the amplifier.

Since you are using an older OptiSystem version, do you have access to OptiSystem sample files that I mentioned in my previous comment? If yes, you can see how the Raman amplifier is pumped for average model. The upper pump frequency can then be set according to the frequency chosen for the pump.

I don’t think it would be of any help if I upload a file showing how it can be done since you are using an older version of OptiSystem software and it is not possible for you to open it. So can you please consider my question in reply #29714 where I have asked about the OptiSystem sample .osd files?

Hi Swapandeep,

You are right. I checked the system and realized what you mentioned is happening. Even I increased sequence length and no difference was made. And this has nothing to do with the optical amplifier being used in the system, since the system had the same problem even by disabling the amplifier component.

Regards

My expectation is that the system performance should always deteriorate as the distance increases. And as I mentioned in my previous comment, I omitted the EDFA amplifier that you were using and still had the same results, with BER being better at longer distances. So this had to do with the fiber channel rather than amplifier.

I cannot find any attached paper to your comment Swapandeep. Please also address the exact diagram after attaching it. I would also like to know about the work done in the thesis which suggests the system performance should become better as the distance increases when self-phase modulation is disabled.

Thank you for the attachments. Sure the diagram shown can be implemented using other amplifier types. However, in many applications the Raman amplifier is used as a preamplifier, meaning it is used before the receiver. But if you want to compare their performance in terms of power booster (which is used after the transmitter), it is O.K.

As I mentioned, it depends on whether you would like to implement power booster, in-line amplifier or preamplifier. I suggest you go with the one you uploaded earlier in case of in-line amplifier or preamplifier. It really depends on the application of the amplifier and you have to determine that first.

I pointed out in our previously discussed issues in the same topic that self-phase modulation and many other nonlinear effects influence the system only when the power of the signal entering optical fiber is too high. Otherwise, their effect can be negligible depending on the level of the optical power.

You’re welcome Swapandeep.

Regards

What I suggested in my former comments was to use laser pump for Raman amplifier in your implementation. After that, you should set the upper pump frequency according to the frequencies you chose for the laser pump. Have you tried using laser pump as I suggested? If so, please upload the file.

I ran your simulation and managed to get results. To improve the performance of the amplifier, you may consider using array pump even though it increases the cost of amplifier. I have attached an example of laser array pump used in your system. You wouldn’t be able to open the .osd file so I have just attached a screen shot.

You’re welcome, Swapandeep. From reply #29786 I thought there may be something wrong with it (for example Q-factor of about 3.5 is inadequate for your purpose), but then I believe comparison can be made with all qualities of signal, taking into consideration the fact that Raman can perform better being given more pumps.

Regards

If you are comparing Raman-SOA with Raman-EDFA, definitely not. Otherwise, if you are for example writing a paper or something, I believe the purpose of your work really matters. You can see first why multi-pumped Raman amplifier is used in the first place to see if it is necessary for your work.

Here is a link from Optiwave where the use of multi-pumped Raman for gain flattening is discussed. You can see a pump array with 12 lasers has been used for this purpose and the references are all mentioned at the buttom:

Flattening the gain of broadband Raman amplifier with multipump configuration

You’re welcome, Swapandeep.

Regards

Since the dual port WDM analyzer calculates the gain and flatness of your optical amplifier, you should connect it across the optical amplifier as you have mentioned. In case you are making use of a hybrid optical amplifier, then this means connecting it across both amplifiers used in the design.

Can you please also address the corresponding figure in the thesis? I could not find it with a look into it (if that is the thesis the graphs should be in). Also it would be helpful if you address the parts in the thesis where the systems settings are determined, the comparison between implementations can be made.

If you intend to get results more or less the same as the ones in the figures, then the system configuration should be more or less the same. Even though this should be done, I still believe that there should be some similarities between the results that are obtained by both system configurations.

Not really. I believe your system configuration is OK, but it is different from that of the systems implemented and examined in the aforementioned thesis. The most prominent difference between your results and those of the thesis is related to the results obtained for Raman amplifier in my opinion.

The trend of changes in the figures you have obtained is very much similar to those in the thesis, so the implementation is OK. I think the differences in values is because of the differences in parameter settings. Unless you use the exact settings, it is hard produce the same results (with the same software). In the former figure, Raman was very different, but now EDFA is different.

You’re welcome Swapandeep. They have proven more effective especially in long-haul systems in terms of gain flatness, gain value, bandwidth and all in all, they are more cost efficient to be used there. However, each type of optical amplifiers have their own advantages and disadvantages and for some other applications, use of other hybrids might be better.

You’re welcome Swapandeep. What is mentioned is the reason behind the gain flatness that I mentioned in my previous comment, which I believe to be one of the most important features of hybrid Raman-EDFA (otherwise high gain can be achieved using other hybrid amplifiers or non-hybrid amplifiers).

There are analytical calculations of BER which are presented in digital telecommunication courses and regard basic fundamental concepts. Are you seeking such explanations? As for the basic concepts of dispersion, i suggest you go through the book “optical fiber communications” by Agrawal, unless you would like specific information.

Indeed the information provided in Mr Agrawal’s book is quite helpful. However, it may take quite some time to study the material. I have gone through the explanations given in the book written by Mr keiser more concise and more suitable for study to achieve general information on different optical amplifiers.

For better understanding dispersion for your purpose, I believe the book written by Mr Agrawal (optical fiber communication systems, almost all of the editions) are suitable, as the book discusses the issue from a very basic point. reading this, the dispersion can be calculated in the optical fiber.

Unless the design of the optical amplifier itself is of importance, I believe black box EDFA component can be used. You only need the gain and noise figure to consider in your calculations as you have pointed out. You have disabled nonlinearities, so you should consider noise figure and dispersion in your calculations.

Could you please explain what you mean by “negative gain”? I have heard of the cases with the gain having values between zero and one in electrical components and therefore causing loss. However, if the amplification is not done in the desired way, appropriate setting of the EDFA amplifier component is required.

To calculate gain and noise figure, the information provided by dual port WDM analyzer can be used. However, in case you are using components acting as the black box (such as EDFA black box), there is the option of choosing the gain as the parameter you can set to be constant in the whole spectrum.

Thank you for the explanation. Negative values in logarithmic scales are the same as the values between zero and one as I mentioned in my previous comment. If you are using the EDFA component (not the black box), you should change the parameters according to the work you are trying to reproduce.

I don’t exactly understand your question regarding how the calculation is done. Maybe it will be helpful if I mention the other notation used for this purpose: A number can be written either as 5e11 or 5*10^11. So 4*10^15 that you have mentioned can be written in the following form also: 4e15.

This can easily be solved using Matlab software. neper number to a certain power can be expressed using exp() function in Matlab, where in the paranthesis the expression for the power should be written. In your case, you would have to write exp(4e15), which yields a very very large number (Inf).