- This topic has 137 replies, 6 voices, and was last updated 8 years, 7 months ago by
Irfan Morshed.
-
AuthorPosts
-
-
October 1, 2015 at 2:45 am #25544
Swapandeep Kaur
Participanti am a beginner. please help me with the basics…
-
October 1, 2015 at 3:31 am #25552
LALIT VERMA
ParticipantWhat type of SAC ocdma u want to implement and what is ur base paper on which u are working b coz sac ocdma is implemented in many ways.
-
October 1, 2015 at 4:55 am #25567
Swapandeep Kaur
ParticipantSir i am implementing the attached paper . please guide..
-
October 1, 2015 at 5:10 am #25570
alistu
ParticipantHi Swapandeep,
Can you please tell me if you have gone through my very basic explanation of the SAC OCDMA in your topic optiwave.com/forums/topic/problem-while-installing-optisystem-13/ . I believe it will help with implementing your attached paper system, as it gives some basic information. Then you can ask about more specific details.
Regards
-
-
October 1, 2015 at 5:23 am #25572
Swapandeep Kaur
ParticipantAlistu sir i have gone through everything you sent in the link but it does not provide the basic explanation and i am not getting how actualy the sac ocdma is being implemented… we are using fbg but where in optisystem is the option that generates the codes like 101011.? This is the first time i am using optisystem.
Regards
-
October 1, 2015 at 5:34 am #25573
Taiwo Ambali Abiola
ParticipantFirst of all, you are welcome to the forum the same way I would equally welcome myself after a long break due to a number of task at hand. Just as Lalit Verma has already said, the forum is more of a way to implement your design using optisystem softwares and some others. There are some factors you need to consider when talking about SAC-OCDMA. One is how many users, which light source do you want to use (LED or multi-wavelenght laser,white light), which code do you plan to use, (eg, OOC, MQC, KS, MS code) and also the decoding techniques you have chosen to use.
Whichever way, You have the light source (example LED), the spliter(divide into number of users), the FBG to implement the chosen code, modulator to modulate the light using the bit sequence from the PRBS with a specific modulation format ( such as NRZ, RZ, and other advance modulation formats such as QPSK, DPSK etc). Then you need a power combiner/coupler to couple the signal from different users. then you select your channel of communication (free space optics (FSO) or fiber). At the receiver, you equally need a splitter, to divide the combined signals into the number of receivers. Then you apply your chosen decoding techniques e.g(direct decoding, complementary, AND decoding or any other you have chosen). after selecting the decoding measure, you need a photodetector (to convert to electrical signal), low pass filter (to remove noise) and also BER analyzer (which gives you error rate by comparing your received signal with the transmitted one, then gives you the rate of error in your system. If you can recollect, A minimum BER of 10^-9 and below e.g (10^-10, 10^-15, 10^-18, 10^-30 etc) is the standard for effective communication. You can modify the parameters to suit your design. I hope the explanation helps. Kindly signify if you need more help. There are professionals like Ravil, Samson, Optiwaves teams, and others to help with more information.
-
October 1, 2015 at 5:46 am #25574
Swapandeep Kaur
ParticipantThank you so much sir for the useful explanation. since i am a beginner , i am not making any changes to the sac-ocdma given in the sample of optisystem 13 and with its help implementing my paper. it involves the use of white light, i want to know why we have used white light?how the fbgs are used to generate the hadamard codes . since the sample involves the use of three users, i too want to implement the same. only after getting what is being done , i can think of changing the users or other parameters….
-
October 1, 2015 at 6:09 am #25578
alistu
ParticipantSo I assume you have understood how the code is actually generated, and you just wish to know how to implement it. Imagine a codeword obtained for a user to be, for example, 101000. As you can see, only the first and the third elements are ‘1’ and the rest are zero. This shows that for this specific user, you have to use two FBGs whose wavelengths are tuned at lambda1 and lambda3 (Why? Because only the first and the third elements were non-zero at 1010000). That’s it!
Now how are lambda1 and lambda3 chosen? For example, you can start from lambda1=1550.1, lambda2=1550.2,… .So lambda7 would be 1550.7. This is arbitrary. You just choose a starting and and step wavelength (And you’d better you LEDs wavelength the middle frequency of the lambda range). Please tell me if this explanation helped.
Regards
-
-
October 1, 2015 at 6:14 am #25579
Swapandeep Kaur
Participantsir this information is really useful. but actually i did not get how codes are generated ,so please help me with that. and why white light is chosen?
-
October 1, 2015 at 7:02 am #25592
alistu
ParticipantWhite light is a combination of different wavelength, and here also we need a spectrum so that some wavelengths would be filtered in it.
I guess now my explanation in the link below:
optiwave.com/forums/topic/sac-ocdma-with-md-code/#post-22736
along with the paper “Development of a new code family based on SAC-OCDMA system with large cardinality for OCDMA network” are going to be very useful in understanding how codes are generated. I have also explained it very simple there, so you will hopefully find that useful too and if not, please let me know.
-
-
October 1, 2015 at 6:18 am #25580
Swapandeep Kaur
Participantsir the concept you just explained i have understood. thank you.
-
October 1, 2015 at 7:55 am #25593
Swapandeep Kaur
Participantthank you sir for the information. but i have to implement hadamard codes in my paper ad in sac -ocdma sample also it is hadamard code not md code
-
October 1, 2015 at 8:07 am #25595
alistu
ParticipantGenerating Walsh-Hadamard codes from Hadamard matrix is much easier. Please look at the following matrix:
1 1
1 0The next Hadamard matrix would be a 4×4 one being generated by putting the above matrix instead of each of the elements of it. But for the zero elements, you would have to change ones to zeros and zeros to ones. Look:
1 1 1 1
1 0 1 0
1 1 0 0
1 0 0 1The above matrice is composed of for matrices as depicted above and shown below:
1 1 – 1 1
1 0 – 1 01 1 – 0 0
1 0 – 0 1I hope the explanation is clear. Please compare the two matrices to find the relation. Now every row or column of the above matrices is a code for one user.
-
-
October 1, 2015 at 8:10 am #25597
Swapandeep Kaur
Participantyes sir it is clear thank you sir
-
October 1, 2015 at 8:41 am #25601
Swapandeep Kaur
Participantsir suppose in the paper its shown at the encoder 11001100 (has that been chosen by choice or how by codind?) it will require 4 fbg. each 1 wil require 1 fbg and correspondingly frequency we will set, that i have understood. but the code at the encoder, how that has been chosen??
-
October 1, 2015 at 9:07 am #25609
alistu
ParticipantThe 11001100 code is the third column of the 8×8 Hadamard code (you could just use the code “hadamard(8)” in Matlab to see this).
I don’t understand your question. Based on the number of users, you can choose hadamard code length (for example for 8 users, you can use each of the rows or columns of the 8in8 hadamard matrice as a code for one user). If this is not your answer, please elaborate more on your question.
-
-
October 1, 2015 at 9:39 am #25620
Swapandeep Kaur
ParticipantThree users are there and the chosen codes at the encoder are 10101010,11001100,10011001 in the paper i am implementing.They have been chosen from the 8×8 hadamard matrix? one is second column,third column and 4th row. Why 8×8 matrix is chosen? and any row and column can be chosen thereafter i think as we want
-
October 1, 2015 at 12:33 pm #25624
alistu
ParticipantThe most probable reason for doing such a thing is better performance I suppose, as in the electrical CDMA (not always, though). And I made an 8in8 matrix for an 8 user system just as an example. But obviously the number of rows or columns cannot be less than the number of users.
All vectors created from matrix rows are orthogonal to each other, hence all vectors from columns. So you can pick any rows or columns.
-
-
October 1, 2015 at 9:40 am #25621
Swapandeep Kaur
ParticipanthadamardMatrix =
1 1 1 1 1 1 1 1
1 -1 1 -1 1 -1 1 -1
1 1 -1 -1 1 1 -1 -1
1 -1 -1 1 1 -1 -1 1
1 1 1 1 -1 -1 -1 -1
1 -1 1 -1 -1 1 -1 1
1 1 -1 -1 -1 -1 1 1
1 -1 -1 1 -1 1 1 -1
will be as above (zeroes instead of -1) -
October 5, 2015 at 1:51 am #25800
Swapandeep Kaur
Participantplease explain the attached ber analyser graph.what does it signify?
-
October 5, 2015 at 2:16 am #25802
alistu
ParticipantHi Swapandeep,
Generally speaking, 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 in the signal path.
– The eye width corresponds to timing synchronization and jitter effects
– Eye closure corresponds to inter-symbol interference in your system.With regards to the above factors, you can identify the destructive effects in your system from the information in the eye pattern.
Regards
-
-
October 5, 2015 at 2:47 am #25808
Swapandeep Kaur
Participantwhat does the blue curve in the figure i have attached signify? its not part of the eye diagram
-
October 5, 2015 at 1:23 pm #25849
Damian Marek
ParticipantThe light blue curve you mentioned plots the Q factor as a function of the decision instant. Normally the Q factor is the highest in the center of the eye diagram which corresponds to the largest different in the signal amplitudes and near the edge of the eye it drops off. By changing the tabs at the bottom of the figure you can plot other results as a function of the decision instant, for example BER.
-
-
October 5, 2015 at 9:25 am #25833
Swapandeep Kaur
Participantwhat does max Q factor , min ber, eye height , threshold signify? whether they should be high or low?
-
October 5, 2015 at 9:43 am #25842
alistu
ParticipantIn order to have a comprehensive understanding of the factors mentioned above, please go through the file attached in my comment given in the below link:
In the meanwhile, let me point out that the more eye height, the more Q-factor and the less BER, hence the signal better. Please read the paper and ask any questions you have.
-
-
October 7, 2015 at 9:48 am #25945
Swapandeep Kaur
ParticipantThank you sir. sir i implemented sac ocdma with long reach passive optical network. i am not getting the output. please guide. the figure was large, so i am attaching the file in two parts.
-
October 7, 2015 at 11:59 am #25953
alistu
ParticipantThe snapshots you have attached help me implement the system and then try to troubleshoot it, but it would be more helpful if you kindly upload your own implementation of the design in optisystem (the corresponding osd file). . Please attach the file and I will look into it. Are you using the latest version?
-
-
October 7, 2015 at 9:52 am #25948
Swapandeep Kaur
Participantsir i used fiber bragg grating. i think i should use uniform fiber grating
-
October 7, 2015 at 10:39 am #25949
Swapandeep Kaur
Participantsir didnt get the output with uniform fiber grating too..
-
October 7, 2015 at 12:23 pm #25954
Swapandeep Kaur
Participantsir these are snapshots of the file that i implemented..what does osd file mean?
-
October 7, 2015 at 12:24 pm #25955
Swapandeep Kaur
Participantyes sir i am using the latest version optisystem 13
-
October 7, 2015 at 12:30 pm #25957
Swapandeep Kaur
Participantosd format
-
October 8, 2015 at 8:32 am #26002
Swapandeep Kaur
Participantsir were you able to find out where I am going wrong?
-
October 8, 2015 at 9:11 am #26005
alistu
ParticipantNot unfortunately, as I am not familiar with the receiver structure and the basis on which the wavelength for each of the fiber bragg gratings for each user is being chosen. I only know you have used hadamard matrix to produce the code words. Can you elaborate on that please?
-
-
October 8, 2015 at 9:24 am #26006
Swapandeep Kaur
Participantsir I have chosen the wavelengths as you had guided and taking the sac-ocdma simulation as the reference. i took 1550.1 nm as the first wavelength and 0.3 nm as the bandwidth and corresponding to that according to the rows and columns of the hadamard code selected in the paper like 10101010 , selected the next wavelength as 1550.7 nm and so on. at the receiver side, the upper decoder wavelengths for code 10101010 were starting from 1550.1nm and the lower decoder as 1548.5 nm
-
October 8, 2015 at 9:44 am #26012
Swapandeep Kaur
Participantsir will this information help?
-
October 8, 2015 at 10:09 am #26016
alistu
ParticipantI hope so. I haven’t got the results yet and I will let you know if I manage to do so. Can you please tell me how you chose the lower decoder wavelength?
By the way, please have a closer look at the bit rate of your channels. You may want to set them again, as one is not set properly and the others are set at the global bit rate (and not at 200Mbps, as in the Optisystem example which you might have been trying to set to).
-
-
October 8, 2015 at 10:02 am #26015
Swapandeep Kaur
Participantsir please reply. this is my college project and i am not getting the output………..
-
October 8, 2015 at 10:15 am #26017
Swapandeep Kaur
Participantsir i chose the lower decoder wavelength with reference to the sac- ocdma sample, 1548.5 as the initial and the next with .3 as the bandwidth and 10101010 code.
-
October 9, 2015 at 10:30 am #26098
Swapandeep Kaur
Participantsir did you get any output?
-
October 9, 2015 at 11:04 am #26099
alistu
ParticipantNot unfortunately. And I am not still sure about the basis on which the wavelengths of the fiber bragg gratings on the receiver side should be chosen. I will let you know if I make any progress. I think members with an experience in this area such as Lalit (with whom we formerly discussed a SAC OCDMA system in another topic) may provide valuable insight.
-
-
October 12, 2015 at 8:22 am #26251
Swapandeep Kaur
ParticipantSir did you get any output? I asked the authors of the paper, they sent a snapshot of the parameters, i am attaching that, see if that helps
Regards
-
October 12, 2015 at 8:46 am #26257
alistu
ParticipantHi Swapandeep, I have done the simulation with a bit rate of less than this in a back-to-back system and have not managed to get the results. I think the information we need is the wavelength of the lasers and the wavelength of FBGs both on transmitter side and on the receiver side, as I am not sure about the theoretical basis of choosing wavelengths in the receiver.
-
-
October 12, 2015 at 10:43 am #26262
Swapandeep Kaur
Participantsir is there nobody who could help us…
-
October 13, 2015 at 9:08 am #26361
Swapandeep Kaur
ParticipantSir in the OCDMA network design I read this line “The diagram below shows the receiver-end for one of the users, where the received signal is split into an upper branch (FBGs are the same as the encoder) and lower branch (FBGs are the complement of the encoder)”.
According to this the wavelength is set i think -
October 13, 2015 at 9:21 am #26363
Swapandeep Kaur
ParticipantSir what i took in my simulation was wrong , since i didn’t know this concept.
-
October 13, 2015 at 9:37 am #26365
alistu
ParticipantYou’re right. And now that I wanted to correct it according to your statement, I realized you have used 1550.1 and 1550.9 while you have taken the steps as 0.1 nm. Therefore your code length must have been longer than 8. If I am wrong, please tell me how you have created the code so that we would be able to correct the system.
-
-
October 13, 2015 at 10:16 am #26384
Swapandeep Kaur
ParticipantSir i constructed the code like this : if code is 10101010 , then if first wavelength is 1550.1 and bandwidth is .3, the next would be 1550.7nm , though know i think that the steps and the bandwidth have no relation with each other ( as seen from the image that i sent yesterday that was sent by the authors of the paper) they have used 1550 , 1550.5 , 1551 ( say for 11110000) while bandwidth is 0.3 nm
-
October 14, 2015 at 9:26 am #26501
alistu
ParticipantYou have named three wavelengths for a code of length 8 and with four ‘1’s, while four FBGs with four different wavelengths should have been used. Are you sure about the wavelengths used in the paper?
I suggest that you design the system with central wavelength of 1550.5 with steps of 0.5nm and bandwidth of 0.3nm and choose the receiver FBG wavelengths as you earlier mentioned (by using complements). If you did not get the answer, please attach the file here.
-
-
October 13, 2015 at 10:56 am #26399
Swapandeep Kaur
Participantsir any hopes of getting result? i am confused…
-
October 13, 2015 at 10:59 am #26401
Swapandeep Kaur
Participanti am confused what will be the codes at decoder if the lower fbg are complement of upper ..
-
October 14, 2015 at 8:59 am #26499
Swapandeep Kaur
Participantsir this is what i got from a thesis work
The fbg cut off freq is chosen randomly within the broadband light source bandwidth. Furthermore, to take in notice on choosing the cut
off frequency is not to choose a cut off frequency that is close to another bit and not choosing the cut off frequency that is outside or too near to the upper and lower side of the broadband lights source bandwidth.
The Specification of the Output from the spectrum is stated below:
White Light Specification
white light source freq 1550.5nm
white light source power -115dbm
fbg cut off freq 1548.5nm,1550.1nm 1550.9nm 1552.5nm
bandwidth .6nm
reflectivity .9998-
October 14, 2015 at 9:29 am #26502
alistu
ParticipantPlease attach the thesis you are referring to, or address it by giving the name, name of the author and name of the university. As for what you mentioned about the bandwidth, I also think 0.3nm is a small amount for the bandwidth. But it would be better if you use script mode to set bandwidths, so that if we have to change it, we wouldn’t have to change them one by one.
-
-
October 14, 2015 at 9:14 am #26500
Swapandeep Kaur
ParticipantIt is given that for smaller bandwidth, there can be more users involved but the ber is high . best results are obtained at bandwidth of 0.6nm
-
October 14, 2015 at 9:48 am #26505
Swapandeep Kaur
Participantthe thesis
-
October 14, 2015 at 10:32 am #26510
alistu
ParticipantThank you for the thesis. It is somehow expected the less bandwidth is used, the more error might be made in the system. So you’d better use the script mode so as to be able to see what difference the change of bandwidth makes in your system. Please also consider my reply #26501 on this topic.
Regards
-
-
October 14, 2015 at 10:48 am #26519
Swapandeep Kaur
Participantyes ill consider your reply . ill chose the central wavelength as 1550.5 and steps as 0.5 nm . because i didnt get what step size has been chosen in the sac- ocdma example.
one more thing i learnt from the thesis was that we can check the code with the spectrum analyser after the encoder..
-
October 14, 2015 at 10:51 am #26522
Swapandeep Kaur
Participantwhat white light freq should i choose? 1550.5 nm? and what about the layout parameters?
-
October 14, 2015 at 10:52 am #26523
Swapandeep Kaur
Participantplease explain how to take complement of the codes at the decoder?
-
October 14, 2015 at 11:22 am #26525
alistu
ParticipantYou can choose white light frequency the same as the center frequency that you have chosen for your codes. The complement codes are formed from the code used at the transmitter when ones turn to zero and zeros turn to one. For example, if your code line is “11110000”, the complement is “00001111”.
-
-
October 14, 2015 at 11:49 am #26529
Swapandeep Kaur
Participantsry sir one more question what does central wavelength mean? it is 8 bit code , so what will be the centre frequency?
-
October 14, 2015 at 12:23 pm #26532
alistu
ParticipantWhat I refer to as the center frequency is the frequency which is in the middle of the frequency range used for the fiber bragg gratings. For example, for a code of length 8 starting from 1550 with steps of 0.5 nm, lambda1 would be 1550 nm and hence lambda8 1553.5 nm. So the middle wavelength is 1551.75 nm.
-
-
October 15, 2015 at 4:02 am #26552
Swapandeep Kaur
Participantsir what if i choose the centre freq as 1550 nm bcoz 1550.5 is getting tough to be chosen as centre freq
-
October 15, 2015 at 4:14 am #26553
Swapandeep Kaur
Participantsir im selecting the freq as 1549 1549.5 1550 1550.5 1551 1551.5 1552 1552.5
-
October 15, 2015 at 4:28 am #26554
Swapandeep Kaur
Participantsir there is something wrong. im not getting correct diagram of spectrum analyser after encoder like of code 10101010 . im getting the result as attached
-
October 15, 2015 at 5:23 am #26556
alistu
ParticipantHi Swapandeep,
While I was trying to open your file, I encountered an error stating the file could not be opened. I have attached the screenshot of the folder here. And i am using the last version of OptiSystem. can you please upload your file again? Thank you.
Regards
-
-
October 15, 2015 at 5:38 am #26562
Swapandeep Kaur
Participanti attached the word file
-
October 15, 2015 at 5:39 am #26564
Swapandeep Kaur
Participantshould i attach the osd file?
-
October 15, 2015 at 6:08 am #26574
Swapandeep Kaur
Participantsir i am attaching the osd file. i am not getting the results, there is some problem with the code generation as the spectrum analyser is not giving correct results of the encoder
-
October 15, 2015 at 6:13 am #26576
Swapandeep Kaur
Participantthe osd file
-
October 15, 2015 at 7:16 am #26582
Swapandeep Kaur
Participantsir did you come to know the problem?
-
October 15, 2015 at 9:39 am #26597
Swapandeep Kaur
Participantsir i made changes in the sac- ocdma example and have got some success in the first code and a very good eye diagram…thanks to you….
looking forward to more of your expert comments…. i am attaching the osd file… though second code at the encoder is not correct -
October 15, 2015 at 9:52 am #26602
Swapandeep Kaur
Participantthe starting wavelength is 1549nm, step size is 0.5nm the first code is 10101010..
second code is 11001100, though im not getting that correct, still working on that… -
October 16, 2015 at 7:01 am #26644
Swapandeep Kaur
ParticipantSir i have got the rest of the codes as well. it depends on how we choose the cut off wavelengths.
Sir i am not understanding how to obtain the results portion of the paper i am implementing. please help me with that. i have made the basic simulation of sac- ocdma using lr-pon. i am attaching the paper. please help.
-
October 16, 2015 at 8:10 am #26652
alistu
ParticipantDid you get the suitable results for the rest of the users in your system as well? If not, please mention the problem you have faced. May I also ask you to upload your implementation file so that I would be able to see the system while considering the paper you have attached? Thank you.
-
-
October 16, 2015 at 8:13 am #26653
Swapandeep Kaur
Participantyes sir i have got the suitable results for the resr of the users as well. i am attaching the osd file
-
October 16, 2015 at 9:03 am #26659
Swapandeep Kaur
Participantsir i am not getting how ber varies with the transmitted power for different values of edfa gain? how are we getting different values of edfa gain ? from which option? and from where we get with and without ase noise? please reply…
-
October 16, 2015 at 11:55 am #26682
alistu
ParticipantYou should put a power meter before the transmission line to measure the transmitted power. Put two power meters before and after the EDFA to measure the component gain (or however you measure it). And then you just set and/or change the parameters to achieve the results. You can use sweep iteration capability to get the results easier. I suggest that you do that.
P.S. For EDFA, you can also use the black box to be able to control gain easily.
-
-
October 17, 2015 at 1:52 am #26705
Swapandeep Kaur
Participantsir if i put the power meter before the optical fiber , i will get a fixed transmitted power. how to get different values of transmitted power versus ber?
-
October 17, 2015 at 1:55 am #26706
Swapandeep Kaur
Participantand we have to calculate ber with and without ase, for that we will enable and disable the option in the photodiode?
-
October 17, 2015 at 3:14 am #26709
alistu
ParticipantThe transmitted power is not the one controlled by the optical amplifier in your design. In order to change the transmitted power, you are going to have to change the power of the white light source lasers. And you can enable or disable the ASE noise by changing the potion in the photodetector.
-
-
October 17, 2015 at 4:13 am #26716
Swapandeep Kaur
Participantsir but we have to see the effects of edfa noise ,ASE ;; Is it same as changing noise of photodiode?
-
October 17, 2015 at 4:32 am #26718
Swapandeep Kaur
Participantsir may i ask you a couple of more questions regarding the project ?? such as why we chosen white light freq as 1550.5 nm? why mach zehnder modulator extinction ratio is 30 db
we took edfa so as to avoid the losses. but in the simulation the Q factor does not improve and ber does not decrease with edfa in comparison to without edfa. so what is the importance of edfa??
my paper is like i earlier said The Effects of EDFA Noise on the Performance of Multi-wavelength OCDM-based Long-Reach Passive Optical Networks-
October 17, 2015 at 7:29 am #26726
alistu
ParticipantIf the wavelength of the white light laser is chosen to be at the center of the range of wavelengths used for the fiber bragg gratings, there would somehow be a symmetry in the way the wavelengths far from the center receive power. So it had better be chosen that way.
The chosen extinction ratio has been chosen as the default value. But you can change it according to your paper (if there is any mention of it).
Don’t you see any difference in the performance by changing the EDFA gain?
-
-
October 17, 2015 at 7:37 am #26727
Swapandeep Kaur
Participantsir how to change the edfa gain? i dont know the concept of black box
-
October 17, 2015 at 8:00 am #26732
alistu
ParticipantChnaging the EDFA gain needs designing it for the gain you want it to have. However, in the EDFA black box, you can simply choose the gain you want EDFA to have and that’s it. In case you want to use it, set the operation mode in the EDFA black box properties to “gain control” and apply the gain value you want.
-
-
October 17, 2015 at 7:47 am #26730
Swapandeep Kaur
Participantsir why fbg reflectivity chosen as .9998? noise dynamic 5 db? prbs generator bit rate as signal bit rate?
-
October 17, 2015 at 8:06 am #26734
alistu
ParticipantThe default value for reflectivity of the fiber bragg gratings is 0.99 in OptiSystem. And noise dynamic default value is 3dB for this component. However, you can change the value of these parameters along with other parameters to any other value you desire according to the paper you work on or so.
-
-
October 17, 2015 at 8:34 am #26738
Swapandeep Kaur
Participantsir you said that the transmitted power is the power from the white light source. it has been chosen as -115 dbm. if i am changing it like to 0 to bring a plot vs ber , the simulation is taking too long,the calculation is not ending..
for the edfa black box, we can also select the option of ase noise, since it has the option .
if i am asked regarding the values of the parameters in the viva which is on monday, i can say that a particular combination of values is chosen to bring about the simulation
-
October 17, 2015 at 9:12 am #26742
alistu
ParticipantI think the transmitted power in the paper you are working on refers to the power that enters the optical fiber (after the power combiner), so you can read the transmitted power by a power meter placed there.
Unfortunately, I did not understand your last sentence and would be grateful if you explain it again.
-
-
October 17, 2015 at 8:55 am #26740
Swapandeep Kaur
ParticipantWhy have we chosen the complement of the wavelengths at the lower decoder at the decoder side?
-
October 17, 2015 at 9:24 am #26743
Swapandeep Kaur
Participanti mean to say like the parameters like reflectivity .9998 are chosen to bring about a particular simulation
why is the complement of wavelengths chosen at the lower decoder on the decoder side.
thank you so much sir for your help….
-
October 17, 2015 at 9:32 am #26744
Swapandeep Kaur
Participantbut sir that was the problem, if i use a power meter for the transmitted power it will give one fixed value of power
-
October 17, 2015 at 9:42 am #26745
alistu
ParticipantUnfortunately, I don’t know much about the theory regarding complementary subtraction detection technique. The fixed power you are referring to would change when you change the power of the white light sources (I suggest you make changes to their power equally). It is not dependent on the EDFA, as EDFA is placed after it.
-
-
October 17, 2015 at 9:50 am #26746
Swapandeep Kaur
Participantno problem sir… thank you so much for your help and guidance…
-
October 17, 2015 at 11:01 am #26751
Swapandeep Kaur
Participantsir by changing the power of white light sources the simulation is not getting compltetd. and on changing gain, the transmitted power is getting changed.
-
October 17, 2015 at 12:55 pm #26757
alistu
ParticipantHow much do you change the power of lasers? If you make a huge increase in the input power, the simulation will take a lot of time to be done. However, you won’t need such an increase, as you are trying to reproduce the results mentioned in the paper. I increased and decreased the power of each laser by 3dB and it was run smoothly.
-
-
October 18, 2015 at 1:12 am #26766
Swapandeep Kaur
Participantthank you sir. sir we chose white light but could we have chosen any other source? what would have been the advantages? and why we chose fbgs?
-
October 18, 2015 at 2:03 am #26767
Swapandeep Kaur
Participantsir i read that ocdma uses broadband light source and not cw laser? what are broadband light sources and why they are used?
-
October 18, 2015 at 3:41 am #26772
alistu
ParticipantAs you can see, this component have one input and two outputs. You set the frequency that you want to filter along with the bandwidth and then that part of the signal would be in one of the outputs, and the other part of the signal outside of the bandwidth would be in the other output. Why do we use it? Because it simply suits our design.
-
-
October 18, 2015 at 2:19 am #26768
Swapandeep Kaur
Participantwhy have we chosen walsh hadamard code?
-
October 18, 2015 at 3:46 am #26774
alistu
ParticipantWalsh-Hadamard codes used for this purpose are orthogonal to each other. It means that the dot product of each two of the codes equals zero. This is the most prominent feature of these codes. I think if you study the theory thoroughly, you will get a better understanding of how this orthogonality comes to use.
-
-
October 18, 2015 at 3:46 am #26773
Swapandeep Kaur
Participantsir you are talking about which compnent?white light has only one output and fbg two inputs ang one output
-
October 18, 2015 at 3:50 am #26775
alistu
ParticipantI was talking about fiber bragg gratings. As for the white light source, it can be used when you need a light source with a wide wavelength spectrum, and although CW lasers have a linewidth, they are supposed to generate one frequency. That’s why white light source is better for this purpose.
-
-
October 18, 2015 at 4:03 am #26776
Swapandeep Kaur
Participantthank you sir. sir last question. what is spectral amplitude coding? i am not getting it. it is an incoherent ocdma. but exactly what is it?
-
October 18, 2015 at 4:28 am #26777
Swapandeep Kaur
ParticipantIn SAC-OCDMA format, frequency components of the signal from a broadband
optical source are encoded by selectively blocking or transmitting them in accordance
with a signature code .sir i got this definition , what will be the signature code in terms of our project???-
October 18, 2015 at 4:40 am #26778
alistu
ParticipantSignature code is a code that is dedicated only to one user, like a signature that belongs only to one person. For example, signature code for your first user is 10101010, which means only this user is using this code and using it at the transmitter and receiver, the user can recover its own data from the received signal (I strongly recommend you to go through some basic papers in this regard to solve all your basic questions).
Regards
-
-
October 18, 2015 at 4:44 am #26779
Swapandeep Kaur
Participantok sir. thank you sir
-
October 18, 2015 at 10:51 am #26793
Swapandeep Kaur
Participantsir there is definitely some problem with the edfa black box. each time i run the simulation ,even with same parameters it gives different results
-
October 18, 2015 at 11:45 am #26798
alistu
ParticipantDid you use the settings that I have addressed in my reply https://optiwave.com/forums/topic/how-to-implement-sac-ocdma/#post-26732 for the EDFA black box (on how to set the operation mode)? If so, can you please attach the OptiSystem file you are currently working on? Thank you.
-
-
October 18, 2015 at 11:25 am #26795
Swapandeep Kaur
Participantsir why does in absence of ASE , BER increase with length?
-
October 18, 2015 at 11:49 am #26799
Swapandeep Kaur
Participantsir i replaced the black box with normal edfa, still the values are not constant. and i am not getting the desired results for with and without ase operations
-
October 19, 2015 at 8:23 am #26828
alistu
ParticipantThank you for the file. there are no power meters or visualizers before or after the EDFA component, so how would you know the EDFA gain changes each time the simulation is run? The EDFA amplification seems to be around 17dB and will not have any significant changes when you repeat the simulation.
-
-
October 19, 2015 at 9:51 am #26836
Swapandeep Kaur
Participantsir my optisystem 13 will expire in 3 days , what should i do?
-
October 19, 2015 at 10:22 am #26838
alistu
ParticipantAs I told you, the EDFA gain is not random the way you have stated in your former post. I simply put two optical power meters before and after the EDFA component and for a couple of times running the simulation, the gain was almost the same. So please connect power meters and proceed with the rest of your work, as there are no worries about this issue.
Regards
-
-
October 19, 2015 at 11:21 am #26841
Swapandeep Kaur
Participantsir its not the gain that is changing but the BER at the output…however i will check again.
my optisystem will expire in 3 days, what should i do?-
October 19, 2015 at 12:42 pm #26845
alistu
ParticipantI increased your sequence length to 4096 and ran the simulation for a few times. The results did not have any significant changes. I suggest you do this. (I reduced the samples per bit parameter to decrease the simulation time, but that causes signal degradation).
-
-
October 20, 2015 at 4:50 am #26872
Swapandeep Kaur
Participantsir can i download optisystem 13 using different id on the same system?
-
October 20, 2015 at 6:30 am #26875
alistu
ParticipantOptiSystem can only be installed on each hard drive one time as far as I know, even if you reinstall your windows. However, the OptiSystem setup you have formerly downloaded can be installed on other computers (provided that OptiSystem had not formerly been installed on them).
-
-
October 21, 2015 at 9:33 am #26941
Swapandeep Kaur
Participantsir is there sac- ocdma example simulation file on optisystem 7?
-
October 21, 2015 at 12:39 pm #26943
alistu
ParticipantUnfortunately, I don’t know whether the SAC-OCDMA example is available in that version. However, as far as I know, all the components necessary to implement your design are available in OptiSystem 7, so I think you would be able to implement it in OptiSystem 7 as well.
Regards
-
-
October 29, 2015 at 6:16 am #27254
Swapandeep Kaur
Participantsir i have implemented a three user system, why is that the ber of the user 2 is lesser than the others?
-
October 30, 2015 at 4:43 am #27272
alistu
ParticipantHi Swapandeep,
Which version of OptiSystem are you currently using? Are you still using version 13.0.3 or have you implemented it with version 7? Please attach your implementation file so that I would be able to check it. If the BER changes significantly with each simulation, simply increase sequence length.
Regards
-
-
October 30, 2015 at 3:49 am #27271
Swapandeep Kaur
Participantsir in the paper its given that ber reduces with increase in length in absence of ase, but there is not a significant change in the values when i am increasing the length . in fact , the results are almost same for with and without ase. for ase , we have to enable and disable the options in the photodetector pin? that is what i have done…
-
October 30, 2015 at 4:51 am #27273
Swapandeep Kaur
Participantsir i am using optisystem 7 and ber is not changing significantly. there is very little change with length. sequence length is 1024. should i increse it?
-
October 30, 2015 at 11:41 pm #27296
alistu
ParticipantNo, I don’t think it is necessary. I thought the sequence length was shorter. In that case, the results might change every time the simulation is done and that’s when the results are not reliable. Even if a short sequence is used, there is no need to change sequence length when the results don’t change.
-
-
October 30, 2015 at 4:53 am #27274
Swapandeep Kaur
Participantosd file
-
October 30, 2015 at 5:48 am #27276
Swapandeep Kaur
Participantsir i am getting this graph for channel 2. i can take it that ber is decreasing with length without ase and increasing for with ase.
Sir what is the reason for the increase and decrease?
-
October 30, 2015 at 11:49 pm #27297
alistu
ParticipantIf you are referring to fig.5 of the paper, the BER does not decrease with the increase in length, but rather stays the same. And I think the figure yielded from your won simulation is pretty much similar to the one in the paper, perhaps except for the fact that the blue curve (belonging to the state with no ASE) should be below the red one.
-
-
November 1, 2015 at 7:58 am #27315
Swapandeep Kaur
Participantsir in this architecture which we have implemented , communication is possible only in downstream direction?
-
November 1, 2015 at 8:08 am #27319
alistu
ParticipantYou can only transmit the downstream data as long as you only implement the downstream part of the system (or network) which you actually have. For upstream, you need to choose the scheme and then if you are willing to use a bidirectional fiber, you need to specify the proper wavelengths for DS and US.
-
-
November 3, 2015 at 8:42 am #27399
Swapandeep Kaur
Participantsir please could you guide about the latest topics in optics for thesis purpose?
-
November 4, 2015 at 1:05 am #27407
alistu
ParticipantHi Swapandeep,
I’d rather we discuss this in a topic related to this subject, or in a new topic so as not to deviate from the main subject of the topic in here. However, Space-Division-Multiplexing is an emerging and very promising research topic. It deals with few mode fibers or few core fibers to increase the channel capacity.
Regards
-
-
April 29, 2016 at 12:58 pm #38635
Irfan Morshed
ParticipantHi everyone,
My project is similar to Miss Swapandeep’s. Its a 3 user OCDMA system, done using walsh coding. It is almost working. But however, its not giving me the desired results. Anybody kind enough to please help me out? Please. I am helpless, and it’s a request.
This is my project link :
-
-
AuthorPosts
- You must be logged in to reply to this topic.