While reading PAM-DMT optical OFDM, I read that we send information on the imaginary part of pulse amplitude modulated signal and set real part zero. But we know that pulse amplitude modulated signal is a real signal. Please help me to understand that, what is happening in PAM-DMT that we are getting a complex signal.
Dear Mohd IMRAN Khan ,
It is so that you prepare the OFDM symbol by base band processing where trhe symbols are synthesized in frequency domain in the I channels and the channels
then every one is converted to FFT and then they summed to each other giving amplitude variations of the OFDM symbols. These symbol wave form are made to amplitude modulate the light intensity after making a dc shift of the OFDM symbols. Originally the the light can be in form of pulses and one amplitude modulate these pulses by the OFDM symbols.
The key point here is that one prepare the symbols waveforms and then let them modulated the current feeding the LED diode.
I would like that you refer to the paper:
Article Hybrid modulation schemes for data transmission improvement ...
Best wishes
@ Mohd Khan: In PAM-DMT, each subcarrier is independently modulated by a real number. Thus, in frequency domain it is a real signal. However, after Fourier transform, the resultant time-domain signal in general is a complex signal which is transmitted in the channel. The reason for the time-domain signal is complex is because the real numbers modulating the subcarriers are independent and therefore are not symmetrical among subcarriers in general.
Shouxing Qu : In your sentence "However, after Fourier transform, the resultant time-domain signal in general is a complex signal which is transmitted in the channel", do you mean Inverse fast Fourier transform(IFFT) instead of Fourier transform?
I am asking this because in OFDM at the transmitting side, we use IFFT.
@Mohd Khan: Yes, more accurately, I should say inverse Fourier transform (IFT) instead of Fourier transform (FT).
Talking about FFT/IFFT, it is just a fast algorithm to implement FT/IFT. No difference regarding the results.
Thank you very much Shouxing Qu for your precious time and feedback. But actually, I still have some doubts. In the literature, they are saying that "make real part zero and forward information on imaginary part only." And they are doing this before the IFFT block. I am attaching two images. As far as I understand this can be done by First performing QAM on the message and then setting real part as zero. This will convert Quadrature amplitude modulated(QAM) signal to Pulse amplitude modulated(PAM). Please confirm this.
@ Mohd Khan: I have to tell you that I didn't get a chance to read the original paper of this topic yet, so I am not sure what they are doing exactly. Thus I cannot say much (If you can forward a copy of the original IEEE paper, I may have a look to know and say more).
I notice that there is a block of "Hermitian Symmetry" before IFFT in the figures you attached. This indicates that the time-domain signal after IFFT should be a real signal. Good luck.
- Badges
- Science topic
- Similar topics
- Higher Education
- Universities