Simple Bi-directional Link

[Shahid]

Im a newcomer to Optisystem. I have made a simple Bi-directional link using GPON Downstream and Upstream wavelength. But when I try to run it then Optisystem becomes unresponsive. Im using Optisystem 7.

Im attaching the project file. Any help will be appreciated.

[Shahid]

Now the simulation is running but i dont get the signal in the return direction. That is after Optical Power Meter 2, I dont get any value for power(-100dBm).
Layout iterations are set to 3. Am i not using the optical delay correctly?
Kindly help to design the bi-directional link.
I have attached the new project file.

Attachments:

1. Project1-GPON_BidirectionalJAREEIE3.osd

[Ashu verma]

Find the corrected file

[Ashu verma]

Hi shahid,
I have redesigned your system file by doing sone cirrections,but couldnt able to upload it.I have uploaded it twice
Any How ,your system having the errors due to some open port components .i suggest you to see the optiwave BPon sample in the samples library.Or you csn visit the link

Broadband Optical System Based on a Passive Optical Network (BPON)

You can design or remodify this system according to your project requirement.Change data rate and you will get the desired results.i have a question why you aee doing this for only single user for bidirectional? You can increase users by setting value at power splitter.
I will tey again to attach modified file if to do so.

Attachments:

1. Project1-GPON_BidirectionalJAREEIE2-2.osd

[Shahid]

Actually Im trying to make a GPON but i started with single user. If OK then I add more.
Got your file will check it and get back to you.
Thanks

[Ashu verma]

Welcome shahid,file is working for both the directions at 2.5 Gbps and you can analyze this system using power meter as you checked before in your file and optical spectrum too.Let me know if you will need any help

[Shahid]

I looked at your attached file and its working fine. Although I have a few questions

1. As there was only one user, do we still need to use (Dynamic Y Select)?
2. How does the buffer selector works.

Regards,
Shahid

[alistu]

Hi Shahid,

The Dynamic Y-select is a dynamic optical switch that also incorporates switching times with a switching time constant. Mathematically, it resembles the charging process of a linear capacitor through a linear resistor. You can refer to the following link for more information:

Dynamic Y-Select [Optical System]

And the buffer selector’s input can be any type: electrical, optical,… and its output type is also not necessarily the same as input type.

[Ashu verma]

Hi Shahid
You have no need to use Dynamic switch as you are using only one Tx.
For buffer working read this

Because the majority of OptiSystem components will not calculate if a signal is not available at the input ports, the Iterations and Initial delay parameter will help the user to avoid a situation in which the simulation will not perform because a component cannot find signals at the input port – and the system will be deadlocked.

The concept of multiple iterations and delays is very confusing, because it does not have a physical meaning. It is a simulation technique and requires an understanding of the calculation scheduler of OptiSystem.
If you need to know more go through this optiwave tutorial

Lesson 5: Bidirectional Simulation — Working with Multiple Iterations

[Shahid]

Now I have added another user(by 1xN bidirectional Splitter and dynamic Y select) but when I evaluate the Switching Event time for 2nd user its value appears equal to First user.
How I could change Switching Event time for the 2nd user/subsequent users after first?

Recent project file is attached.

Attachments:

1. Project1-GPON_Bidirectional_2xONU.osd

[alistu]

Hi Shahid, It is interesting for me to know what switching event time is and how you evaluated it. I would be grateful if you help me understand this, by naming the component you made the measurement with and the way you used to measure this factor. Tell me this and maybe I will be able to help. Thank you in advance.

[Ashu verma]

Hello Shaid
Sorry for late reply,I coulnt get time today for forum discussion.Anuhow the switching time depends upon the following parameters if you are using script mode
1.Time slot
2.Seq. length
3.No of users or upstream Tx you are considering
In case if you are using say n users for upstream then ,you can calculate y select1 and 2 as
Dynamic select 1=TimeSlot * (1/Bitrate) *Sequence length / n

Dynamic select2=TimeSlot * (1/Bit rate) *Sequence length / n + Time window/n
Hope i have solved your all questions

[Shahid]

This is from BPON example (link provided by Sam Sung post23064).

[alistu]

I know Shahid, but can you please tell me the “time switching event” parameter and how I would be able to measure it? So that I can check it for you and help you with it if it was possible. I have downloaded your attached file and just need to know what I asked you in my previous reply. Thank you.

[Ashu verma]

Glad to know that system i have attached working good for you.However for Time Dynamic switch values for example 16 user i have attached
a table for you
E.g.

(BitRate 10.0E+9) (sequenceLength 256) (Time Window 25.6E-9)
Find the attachment

Attachments:

1. 

   Capture1.jpg
   

[alistu]

Well, you didn’t tell me about it but if the parameter that you are referring to is something like a delay caused after the signal passing the optical fiber, then it is no wonder why it has the same amount for both users, as both users’ signals are being passed through the same channel with the same wavelength.

[Shahid]

Actually what I was trying to say is that when I hit the evaluate button for script (in component properties—–>Main Tab—-> Swithing event time) the result evaluated for BPON example is different than mine although I use the same script as in BPON example.

Sorry for the late reply. I hope its the answer u were looking for(Im fairly new to Optisysetm environment).

[alistu]

Thank you Shahid. The reason why the switching time event has the same value for both channels is that in all Y-select components, it is defined in script mode as:

TimeSlot * (1/Bit rate) *Sequence length / 8 + Time window/8

Now as you can see, it depends on four parameters which are the same for both users and are set in global parameters (except for TimeSlot, I think).

[Ashu verma]

Hi Alisthu,You have Copied the correct script from the optiwave BPON example,but Shaid talking about the swithcing timings for two users.The script ypu introduced is for 8 users ,that is why the dynamic y select 1 operates on TimeSlot * (1/Bit rate) *Sequence length / 8 and the next dynamic y select 2 use the value TimeSlot * (1/Bit rate) *Sequence length / 8 + Time window/8.Hope you got my point

[alistu]

Hi Sam Sung, That is right. I have mentioned in my comment that the line was exactly the script used there. Shahid was wondering why the two channels had the same value for the switching time event and I simply told him the reason and the fact that he would be able to change the script in order to get the proper value.

[Ashu verma]

Thanks Alisthu for your reply.Its a pleasure to share and discuss the views and concept with you,every time we get to know very new things.Keep it up

[alistu]

You’re welcome Sam Sung. And thank you for being there to discuss the points being raised by forum members and trying to be helpful. As for this discussion, I think you really helped Shahid by correcting the mistakes he had in system implementation and some of your points were actually interesting for me too.

[Shahid]

Thank you Sam Sung and Alitsu. U guys have been of great help. Im very encouraged by your input and suggestions. This has set me on a very exciting journey(simulating optical systems). Thanks again for all your help.

[alistu]

You’re welcome Shahid! If you ever happen to have any question regarding simulation with Optisystem or any other question about optical fiber communication, please don’t hesitate to ask here. And I hope you would also share your knowledge with the other members in these forums as well as using their advice and experience.

Cheers!

[Ashu verma]

Welcome shaid,as Alisthu said,share your queries and we will discuss here to get a fruitful result.i have one question about the system you are simulating that you have took a value of 2.5 G for downstream as well as upstream.As far as i know upstream in GPON is of 1.25 G? Increased data rate effects you r users supported.

[alistu]

Hi Sam Sung, First of all, thank you for trying to help Shahid. You have made a very good point here. I think Shahid should be aware that in passive optical networks, upstream and downstream do not have the same bit rate. So if he wishes to improve bit rate and not stick to the standards, he should increase the downstream bit rate as well.

[Shahid]

Thanks for pointing out. I will make the correction.

[Ashu verma]

Thank you Alisthi,but here i just wanted to know ,either shaid aware to change the bit rate for upstream and downstream differently or he is just using script mode on both sides.i think this is clear from his question.

[Shahid]

Now I have changed the Upstream rate to 1.25Gbps. Now what should be the bit in (Layout Parameters—>Simulation—->bit rate) 2.5G or 1.25G?

[alistu]

Hi Shahid, Now that you want to use different bit rates for the upstream and downstram, you can set the bit rate in the global parameters to 2.5 Gbps and then go to the second tab (PRBS) in the WDM Transmitter, change the value of Bit rate to “Bit rate/2” in script mode. This way, the upstream bit rate will be 2.5/2=1.25Gbps and whenever you change the downstream Bit rate, the upstream bit rate will automatically be set as half the value of the downstream.

[Shahid]

Thanks for the reply Alitsu and Sam Sung.I have made the changes as suggested. But when I check the BER Analyzer(at OLT side) there is no output. But when I change the upstream rate back to 2.5Gbps then the output is there.

I have attached the project file.

Attachments:

1. Project1-GPON_Bidirectional_2xONU2.osd

[alistu]

HI Shahid, Unfortunately I have not worked much with bidirectional PONs and I do not have much experience in that regard to help, but I think maybe when the upstream and downstream vectors are not of the same size, there will occur a problem with the time vectors. I will try my best and will let you know.

[Ashu verma]

Hi shaid,I have pounted out the upstream bit rate according to ITU standards for GPON.However you run your system on 2.5 Gbps ,no issues in that.Even you go for NGPON1 with more data rate .If you want to change the upstream data rate go to PRBS ,set it to normal mode and then enter value 1.25 G.For downstream it is same 2.5G.How many nom of users you have splitted in system? For how much length?

[Ashu verma]

Hi Shaid
I will check your file ,This a common problem in setting the bit rate differnt than Bit rate mentioned in global window.Even this happened to me many times.After modification i will let you knoe,what exactly the problem is.

[alistu]

Found the problem! Well, not exactly but I found the way to skip the problem. Put the global Bit rate parameter equal to 1.25GHz and in the script, put the bit rate for the to ONUs equal to the global bit rate. Put the PRBS for downstream “2*Bit rate”. Some properties of the upstream must match global properties I guess.

[Shahid]

Dear Alistu, done what u said and now have output at OLT side. But the eye pattern at the ONU side has changed. Can u take a look(recent project file attached) and tell me if its acceptable. Also how we can check the Eye pattern ourself? any tips.

Thanks in advance

Attachments:

1. Project1-GPON_Bidirectional_2xONU1.osd

[alistu]

Hi Shahid, Right now I don’t have access to Optisystem software but I will check it and let you know as soon as possible. However, I have a question regarding your comment. May I ask if the both Optical Network Units give the same eye pattern? Somehow a problem I think you made a mentioned of before. Thank you.

[Shahid]

yes both ONU’s have same Eye Pattern.

[alistu]

Thank you Shahid. I ran the simulation and got the eye diagram patterns. Obviously the eye is not wide open yet it shows that interference is not so much that you wouldn’t be able to distinguish the eye. I think it depends on the performance that you are expecting from your system to have to be considered acceptable.

[alistu]

Here is the way to interpret the eye diagram:

– The eye opening corresponds to the additive noise in the signal.

– The eye overshoot or undershoot correspond to the peak distortion the signal sees in its path.

– The eye width corresponds to timing synchronization and jitter effects

– And finally, eye closure corresponds to inter-symbol interference in the system.

Using these, you can understand the destructive effects in your system by looking at the eye patterns. I hope this is helpful.

[Shahid]

Have attached a simple WDM-PON simulation with 2xONU. Kindly provide your feedback on the network design and parameters(insertion loss for components/devices, sampling rate of the receivers etc).

Attachments:

1. WDM_PON_2xONU.osd

[alistu]

Hi Shahid, I ran your simulation and it went smoothly. The results show that the system has been implemented correctly. But I noticed one thing that we have been discussing here about that you had not taken into consideration. The upstream and downstream here have the same bit rate, whereas in real networks, the upstream has lower bit rate.

[Ashu verma]

Hello Shaid,As Alisthu said your system file is correct,you can now increase the users if you want.2.5Gbps is not a issue with this system.However for the more practical signal you can use a WDm multiplexer rather than the ideal one.You can set the extinction ratio of the CW transmitter to 30 as EML.ER 10 refers to DML direct modulatated laser.Same advice for the lazers for upstream also.

[Ashu verma]

OOPS ,By observing your file with a great care i found htat you have implemented WDM -PON not TDM for downstream.Shaid your system is not correct from the view of WDm PON,but correct in terms of ey diagram.You have used 2 different wavelengths for downstream and another 2 for upstream.But you havent demultiplex any wavelength at receiver side of both the cases.So how would you come to know which wavelength you are observing.It is not oblivious that your first wavelength signal go to the first port of power splitter.Same in the case of upstream.You have to choose the correct demultiplexing technique for each certain wavelength.Mow you are receiving combined signal at each eye diagram.

[Ashu verma]

I have found the correct way to implement this system.Use AWG for the routing od the each particular wavelength to particular port.AWG just provide you different wavelengths at different output ports attache to the input port collectively.
Basically AWG is a component or devise used to combine or saprate the signals with different wavelengths.For better understanding you can go through the link i have mentioned below
http://www.rp-photonics.com/arrayed_waveguide_gratings.html

WDM Components – AWG Demultiplexer

[Ashu verma]

Here is a example of solution of everything by optiwave samples
https://optiwave.com/download-1/wdm-pon/
You canalso find this system in optisyste 13.If you dont have then v13 then let me know.AWG is there in optisyste 7 and you can utilize that.For further help ,stay active and take a look at ny all replies step by step.Hope you will understand it Thank you

[Shahid]

Dear Sam Sung, Thanks for your detailed reply. I will try to implement your suggestion but I have few things to clarify:

1) In the Receivers Properties—>Downsampling Tab I have selected center frequency instead of Center Power. Will it not help to receive the correct wavelength?
2) Or I need to place a Bandpass filter before the Photodetector if I want to use a splitter instead of AWG DeMux?

3) If I use the Transmitter as EML(External Modulated Laser), Then I need to use PRBS Generator+Mach Zehnder Modulator with CW Transmitter?
4) What value of Insertion Loss will be realistic for SIMULATION Purpose for following devices, (a) WDM Mux/Demux (b)Bidirectional Circulator (c) Bidirectional Splitter (d) AWG

[alistu]

Hi Shahid, As for what you have asked about the use of external modulators, you are right. You need to use the bit sequence (or the symbol sequence) that contains your data (or coded data) and the CW laser as the optical source. You can easily refer to some of the Optisystem samples to see how this is implemented.

[Ashu verma]

Hello Shaid
1.) You have selected the center frew. it is ok,But as far as i know you have to insert the optical filter to stand out the different wavelengths.By only selecting center frew. does not eliminate the requirement of optical filters and AWG.for example optiwave sample of wdm pon they have used AWG due to reason of routing the wavelenghts to diffrent pon.

For EML you need to use PRBS Generator+Mach Zehnder Modulator with CW Transmitter.For circulators and couplers 3dB and mux demux 3-5dB.

[alistu]

As for your question about the insertion loss, I think 4-5 dB is in many cases realistic for them. However, in many research papers the ideal MUX has been used and this does not prevent generating results and making the point you want to with the simulation. In your case, it can depend on whether you are reproducing a paper and what type of MUX has been used in that paper.

[Shahid]

Hi Alistu, Hope u r having a good day. The design is not from a research paper. Its my first attempt at making a WDM-PON for learning purpose.
Could u also comment on the first two points in my Post#23510.

[alistu]

Thanks. Hope you are having a good day too! Could you please upload the last optisystem file in which all the recent changes have been made so that I would be able to understand your questions more clearly? As for what you have asked about AWG, power splitter functions differently from mux or demux, so it is not like you can use either this or that in a design. Let me see the design and I will provide a more exact answer and also comment on the other point. Thanks.

[Shahid]

Here the Updated file is attached.
I have replaced the Ideal Mux/DeMux with WDM Mux/Demux, CW Transmitter is unchanged from last time(will change it later to EML).
Used 3dB Insertion loss for WDM Mux/DeMux, Bi-directional Splitter. 1dB for Bi-directional Circulators.

Attachments:

1. WDM_PON_2xONU_Corrections.osd

[alistu]

Thanks for the attachment and the details provided about the updates. I do not think it is a good idea to use splitter instead of mux/demux of any type. Power splitter splits the power into a determined number and as a result, the overall power is divided between all the branches. But MUX gives allocates each wavelength channel to a specific output port.

[Ashu verma]

Hello Shaid,I have seen your system and just replaced a photo detector with default one .I have used a besel filter selecting one wasvelength 1546 and i observed better performance of the system .BER resuced to 10^-16 to 10^-18-19 .So i would like you to use optical filter .Rest your system is good.You may increase no. of users by splitting each wavelength 1:n for total supported users.

[Shahid]

Thanks for your comments. I added the Bessel Filter before Photodetector and there is improvement in both Q-Factor and BER.
Do you mean APD when you say default photo-detector?

[Ashu verma]

Hi Shaid
I just replaced PIN with another PIN detector from receiver linbrary not APD.I was in hurry so i coudnt changed the down sampling from you detector.You can use that too.APD is another PD,if you use that may you get better results but APD are costlier than PIN.So yu have to take care of that.Nut it is no necessary that you will get better results all the time with APD.

[alistu]

The use of optical filter is indeed an effective way to boost optical signal to noise ratio. However, I think you can also use the filter at the end of the optical fiber line. I have seen the filter being used this way in some samples and I think this way you only need to use one filter instead of multiple filters.

[Ashu verma]

The MAin use of optical filtres here is to select the wavelength of specific channel which is observed at ONU.In case of Shaid,4 wavelengths has been used 2 for US and another 2 for DS.Now the question is how different wavelengths demultiplexed from a combined data stream.One way is to use AWG for routing.Another is placing 4 optical filters with the specific freq. corresponding to the transmitter side as we do in WDM demux.Here function of optical filter is not to filter noise or passing the the band.

[Ashu verma]

Shaid asked above about the use of peak frequency at the photodetector whether it is PIN or APD.As far as i know you can only get the desired stream of required wavelength by filtering it out.If you selected the one wavelength in PIN down sampling even though you will get the some part of other wavelength too.This is the reason of the boosting in output when we use Bessel filter for specific wavelength.

[Shahid]

Thanks for all your replies. They were very helpful in clearing many of my questions regarding the simulations.
I really appreciate both of you guys(Alistu and Sam Sung) for taking out time to help a junior forum member.
I will implement your suggestion and apprise you with the results.

[Shahid]

I changed the CW Transmitter ER from 10 to 30(without using PRBS Gen and MZM). Is it OK to use the CW Transmitter like that?
There is improvement in BER and Q-Factor after this change.

[alistu]

Extinction ratio determines the ratio of the powers of the high (or logical “1”) and low (or logical “0”) bits used in the transmission. Ideally, we would like the low rate to have zero power but this is not possible and it has an amount. The less the power of zero, the more Extinction ratio and better system performance. This is the reason why you have gotten better results by increasing ER.

[Ashu verma]

As Abhishek said it is ok to use 30dB ER on WDM tx.I think now you need to increase users as you have done this system correctly.

[alistu]

You can either do as suggested or use “power penalty” for compensation. For low values of extinction ratio, the power penalty value has greater changes with the change in the extinction ratio. But in your case, I guess it will be 1 dB at maximum (instead of increasing ER from 10 to 30). So I’m guessing the change in the result has not been that great. Am I right Shahid?

[Abhishek Shrama]

Hello Shaid,Yes you can use. Alisthu question is here for the use of CW laser without external prbs and line coding.I think ER defination we can find on google easily

[Damian Marek]

Hi all,

This topic is being closed as the main question has been answered. If the original poster has more questions please start a new discussion with the proper title and explanation.

Regards

[Author]

Posts