Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Guard Subcarriers in OFDM Useless

Status
Not open for further replies.

rfsystem

Advanced Member level 3
Joined
Feb 25, 2002
Messages
908
Helped
149
Reputation
294
Reaction score
38
Trophy points
1,308
Location
Germany
Activity points
9,548
The guard carriers seems to be useless to me!

1. ISI and symbol timing inaccuracy tolerance is implemented by the cyclic symbol extension.

2. Channel separation is done by possible/certain oversampled filtering after analog filtering.

3. Channel separation in WiMAX is today less than the symbol sampling frequency.

But there are strong reasons to have a 2^k usage of all subcarriers

1. There are better, less PAPR, preamables avaible for 2^k

2. Channel segmentation could be done on binary trees

Is the guard an ugly historical reason? Or am I missing something?
 

if by "Guard interval" you mean the "Zero Subcarriers" in OFDM symbols , i should notice that the zero padding makes the filtering requirements easier to be satisfied in digital domain(if we don't zero pad the symbols , we would need a LPF with a very sharp transition band, which is difficult to implement digitally).

as you said in your post , another way for channel seperation is to use oversampling.

regards
 

Yes, typical the channel filtering requires minimum oversampling of 2x of the symbol sampling frequency to get more relaxed requirements for alias and far away blocking filtering. But exact this reason also state that guard carriers are not sufficient for channel filtering alone. So there is NO specific reason to use guards if oversampled filtering is required anyway.

It remains: What is the true reason for guard carriers?
 

After the IFFT part, windowing(filtering) is done. But the filter is not having a sharp cutt-off frequency. It will have some roll-off factor if the frequencies used in the OFDMA symbol falls in this region, then it will be attenuated. There fore they leave guard band in left as well as right side.
Refer to the attachment if you have doubt on where windowing is done.
 

So for OFDM transmission the signal have to be send to the DAC's. This DAC's operate at 2-8x symbol sampling frequency. Then there is analog aliasing. The interpolation filtering reduce the images in between fs/2 and fdac/2.

What is the net benefit of a little, small guard band of subcarriers, for the effort of interpolation and analog filtering in the transmitt path?

Is the 2^k of used OFDM carriers a better approach for OFDMA than having a little trade gain because of the guards?
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top