DSPRelated.com
Forums

OFDM Block-type Pilot, channel estimation

Started by ezain January 17, 2008
Hi,

I am working on channel estimation in OFDM systems, and have a question
related to block-type pilot arrangement. Block-type pilot arrangement is
one where all subcarriers in an OFDM symbol carry pilots. Does this mean
that this OFDM symbol contains no null subcarriers, even at DC? 

Or is it assumed that these 'pilot' carrying OFDM symbols contain some
minimum number of null-subcarriers at DC and band-edges just like normal
data-carrying OFDM symbols. 

Any help is appreciated. 
Thank you. 

Ekta 



On Jan 17, 5:49 pm, "ezain" <jaine...@gmail.com> wrote:
> Hi, > > I am working on channel estimation in OFDM systems, and have a question > related to block-type pilot arrangement. Block-type pilot arrangement is > one where all subcarriers in an OFDM symbol carry pilots. Does this mean > that this OFDM symbol contains no null subcarriers, even at DC? > > Or is it assumed that these 'pilot' carrying OFDM symbols contain some > minimum number of null-subcarriers at DC and band-edges just like normal > data-carrying OFDM symbols. > > Any help is appreciated. > Thank you. > > Ekta
If the band-edge tones are turned off to satisfy a spectral mask, it would seem like they should be off in the pilots too. If the DC tones are turned off because they lie beneath a DC offset in the receiver, that will be problem for the pilot too. John
Thanks for your reply John. 

Ekta 


>On Jan 17, 5:49 pm, "ezain" <jaine...@gmail.com> wrote: >> Hi, >> >> I am working on channel estimation in OFDM systems, and have a
question
>> related to block-type pilot arrangement. Block-type pilot arrangement
is
>> one where all subcarriers in an OFDM symbol carry pilots. Does this
mean
>> that this OFDM symbol contains no null subcarriers, even at DC? >> >> Or is it assumed that these 'pilot' carrying OFDM symbols contain some >> minimum number of null-subcarriers at DC and band-edges just like
normal
>> data-carrying OFDM symbols. >> >> Any help is appreciated. >> Thank you. >> >> Ekta > >If the band-edge tones are turned off to satisfy a spectral mask, it >would seem like they should be off in the pilots too. If the DC tones >are turned off because they lie beneath a DC offset in the receiver, >that will be problem for the pilot too. > >John >