- This topic has 18 replies, 4 voices, and was last updated 9 years, 4 months ago by
Damian Marek.
-
AuthorPosts
-
-
August 10, 2015 at 11:58 am #23348
NasirParticipantI implement a good VLC system using OFDM with QAM and OOK. Unfortunately, when I run the simulation, it keep showing BER of 0.523 . Of course this is not acceptable in any communication system. Can you Please help me to solve the problem 🙂
-
August 11, 2015 at 10:25 am #23466
Ashu vermaParticipantHello NAsir
According to your question you have implemented a VLC system with different formats and couldnt get a Acceptable BER.For the case of OFDm have you used BER analyzer or BER DUT set,Because it is recommended to use BER DUT set instead of BER analyzer for OFDM due to large sequence length.As far as i know that you have implemented system for 4-QAM ,16-QAm and 256 QAM.Then in general with the increase of the bits per symbol,noise increase and effects of non-linarities for changing the phase.Use of non linearity compensate is good for tht problem.-
August 12, 2015 at 1:28 am #23518
alistuParticipantHi Sam Sung, You have implied in your comment that the use of BER DUT Set is due to large sequence in OFDM systems. Can you please specify why a large sequence is needed in OFDM? I have been working with OFDM and sometimes I use sequences as short as 2000 bits and I am curious to know if any problem happens then. Thank you.
-
August 12, 2015 at 5:05 am #23544Ashu vermaParticipant
Hello Alisthu
As far as i observed,I havnt seen long sequence lengths in OOK schemes where in ofdm system it is 2049,8132 etc and in OOK it is 128,64 etc.I have read in one of th forum post where Damian mentioned the use of BER test set and why we dont get eye diagram as in OOK scheme is due to the large sequence length and low sample per bit say 2 or 4.For the case of higher modulation formats the seq. length is long as compared to OOK.In many posts this has been discussed that BEr test set is preferred over Ber analyzer.
-
-
-
August 11, 2015 at 10:30 am #23467Ashu vermaParticipant
I have just a little question,apart from BAUD rate formats,have you implemented this system for NRZ line coding.You have mentioned about OOK.So you faced same problem with OOK in terms of BER? Or its working fine?Because it is easy if you start with OOK and then go for OFDm and QAM,finding problem in ook system easy due to the clear eye diagram and to conclude from there about different parameters.
-
August 11, 2015 at 10:44 am #23471NasirParticipant
Actually, I used BER Test Set
Please check the screenshots.-
August 12, 2015 at 1:17 am #23517alistuParticipant
Hi Nasir, I see OFDM v.12 component in your design. I remember the use of OFDM in version 12 causing problem for some Optisystem users before. I am not sure if that is the case with the use of the v12 component in optisystem 13, but if you have access to optisystem 13, why don’t you use the new OFDM component?
-
-
August 11, 2015 at 10:48 am #23478NasirParticipant
Yes I faced the same problem with OOK. Actually, I got BER 0.028; which is still not good enough
Please check the attached file for OOK screenshot -
August 11, 2015 at 11:04 am #23481Ashu vermaParticipant
Hello Nasir for the case of OOK as you attached screen shot above, i would recommend you to use low pass bessel filter electrical after PIN photo detector. Instead of attaching BER visualizer to PRBS and RZ you can use 3R re-generator from receiver library. I would suggest you to first see the eye-diagram just after the transmitter to make sure that it is working fine or not? after that check your system to decrease value of FSO attenuation to minimum level for checking. Can you please tell what sequence length you have considered for OOK because in case of OOK sequence length is not large as in case OFDM.
-
August 11, 2015 at 12:49 pm #23490NasirParticipant
Hello Sam. For OOK, I did exactly what you asked me to do. Thank you very much. I think now it is working properly 🙂
Please check the attachment for this discussion and update me if you realize anything wrong. Thank you again-
August 12, 2015 at 5:10 am #23546Ashu vermaParticipant
Welcome Nasir
I am glad that my suggestion helped you.You are working on 20mbps ?Eye diagram you got is seems fine with acceptable range of BER.You can reduce sample per bit to low value for fast calculations.Fo QPSK and OFDM follow the instruction given by Alisthu because i dont have access to latast version.As soon as i get the latest version i will try to help you out.Best of luck
-
-
August 11, 2015 at 12:55 pm #23496NasirParticipant
for OFDM-QAM part, is there anyone to help????
-
August 12, 2015 at 1:13 am #23516alistuParticipant
Hi Nasir, I have noticed you have been using quadrature modulator in your transmitter, but you are not using quadrature demodulator on your receiver side. Now unless this has been done on a purpose, I strongly suggest that you use a quadrature demodulator on the receiver side as well, even though what you have done now seems to be correct.
-
August 12, 2015 at 5:59 am #23553NasirParticipant
Hi alistu. Actually, I used the blocks diagram for Quadrature demodulator on receiver side. It can be found in the attached picture below. but never-mind I will change it
-
August 12, 2015 at 6:11 am #23558alistuParticipant
Thank you Nasir. I am not saying this would definitely solve the problem but at least it might take our concentration on one issue and enable us to concentrate on other stuff that might be causing the problem. there have been some cases in which there was a problem in the system and just by implementing the same system again with minor differences, the problem has been solved! You can find some OFDM topics regarding this in the forums.
-
-
-
August 12, 2015 at 2:01 am #23521alistuParticipant
In the case of the coding you are using, the large number of constellation symbols implies that your system might be very vulnerable to noise. However, I ran your system back-to back and still got 0.5 BER, which tells that there is probably an implementation problem. So please do as I asked you to and let me know of the result. thank you.
-
August 12, 2015 at 10:31 am #23596NasirParticipant
Hi alistu, now I am using Quadrature demodulator. and about constellation, I used four iterations with different BitsPerSymbol, 2, 4, 6, and 8 respectively. so 4QAM, 16 QAM, 64 QAM, and 256 QAM are used in the model.
Actually, I am still facing the same problem.
Please check the attached file.-
August 12, 2015 at 11:38 am #23599alistuParticipant
Hi Nasir, I am sorry I also manipulated the system but could not obtain the correct results. May I ask what version of Optisystem you are using? Just know that the problem with OFDM components in version 12 is a very common one and has been discussed many times in the forums. It’s not bad to have a look at them.
-
August 17, 2015 at 9:52 am #23762
Damian Marek
ParticipantI am not familiar with this type of OFDM. You have 60 LEDs all being modulated by the same signal operating at the same optical frequency. Can you explain your desired system in a little more detail? I would suggest starting with a simpler design maybe one LED and see if you can get the hang of it.
Also there is a quick fix you can do by setting proper settings in a Decision component for the old OFDM components. Follow the example at:
Regards
-
-
September 24, 2018 at 7:02 am #49532
R.Sai SarathParticipantSir we are trying to implement the ook with the same parameters what you have mentioned. But still we are not getting the exact eye diagram. Here im attaching the eye diagram output and the layout paramaters.Once can u see and reply it as soon as possible.
-
-
AuthorPosts
- You must be logged in to reply to this topic.