What type of fadign does the OFDM system suffers?

[malaylah]

Questions about OFDM.

What type of fadign does the OFDM system suffers? Rayleigh Flat Fading or Frequency Selective Fading? Can anyone help me on this? Or any related paper?

 

[mimomod]

Questions about OFDM.

Hi malaylah,

OFDM is a transmission techniques to avoid frequency selective fading. So, it is basically used in frequency selective fading channels, otherwise you don't need OFDM.

Now, OFDM will transform a frequency selective fading channel into multiple (paralel) flat fading channels. The basic idea is to transmit multiple paralel information data with multiple sub-carrier which are orthogonal to each other.

There are plenty of papers discussing OFDM.
I recommend you to read this book , instead.

best

 

[aelmh]

Questions about OFDM.

use this website as well my friends it 'll be helpful www.ebooksclub.org

 

[moustafahussein]

Re: Questions about OFDM.

It suffers from both, as OFDM was created to over the raylieh fading and frequency selective fading.
The concept of sending data parallel to each other at lower data rate which over give a high data rate gets over raylieh fading while the orthogonality of the carriers gets over the frequence selective fading

 

[malaylah]

Re: Questions about OFDM.

So that means.....OFDM is used in frequency selective fading channel.....but...actually.....each symbol from the subcarrier suffers flat fading....am I right? Therefore, to model it during simulation......received signal for each subcarrier is equivalent to:
Let say this is the 1st subcarrier:

received_signal=rayleigh_flat_fading_variable*transmitted_signal+gaussian_noise

And the rayleigh_flat_fading_variable .....I can generate easily with a very basic rayleigh distribution theorem.....which is without consideration of doppler frequency and so on.......Am I right? Pls clarify.....

 

[mimomod]

Re: Questions about OFDM.

received_signal=rayleigh_flat_fading_variable*transmitted_signal+gaussian_noise
And the rayleigh_flat_fading_variable .....I can generate easily with a very basic rayleigh distribution theorem.....

Yes, right.

which is without consideration of doppler frequency and so on.......Am I right? Pls clarify.....

This is not right. Even flat fading channel may experience Doppler effect.
The Doppler PSD (power spectral density) captures the time variation of the channel.

See "Mobile Fading Channels" by Matthias Paetzold for further explaination. The book has been uploaded in this forum.

best

Added after 22 minutes:

It suffers from both, as OFDM was created to over the raylieh fading and frequency selective fading.

To summarize my knowledge on fading channel:
1) Channels can be divided into flat fading and frequency selective fading channels. Flat fading channel will be experienced by communication signal having symbol period larger than delay spread, i.e. the maximum delay of paths taken by signal from the sender to the receiver. If the symbol period is shorter than the delay spread, then the signal experiences frequency selective fading which causes intersymbol interference (ISI). Thus, flat fading channel has only one path and frequency selective fading channel has multiple paths.

2) From the experiments, each path of fading channel has a certain distribution, i.e. either Rayleigh (if no line of sight), Rice (if there is line of sight), or Nakagami (more general). For example in the rural area (RA), the first path (lowest delay) of a frequency selective fading channel has the Rice distribution, whereas the other paths have the Rayleigh distribution.

3) Furthermore, due to the movement of the sender or receiver, there will be time variation of the channel due to Doppler effect. This Doppler effect will be zero only when there is no movement of the sender/receiver.

So the type of a wireless channel could be the combination of three items discussed above. Examples:
a) flat fading channel with Rayleigh distribution and no Doppler effect
b) frequency selective fading channel with all paths having Rayleigh distribution with Doppler effect.
c) etc.....

best