Forums

Choose equalizer or channel coding?

Started by mguo March 4, 2006
This is a very nice forum and a lot of knowledgeable people here. I have a
question don't know if any experts can share some thoughts on.

If you are using GMSK over a wireless channel, and if you can choose only
one feature to improve your system sensitivety, will you choose adaptive
equalizer or channel coding? Which provides more benefit? 

GSM has both while DECT seems to have none.


in my opinion channel coding will be better. but it depends on your
work . how much is realistic your work in terms of your application. do
u familar space time coding ...... u can try !

mguo wrote:
> This is a very nice forum and a lot of knowledgeable people here. I have a > question don't know if any experts can share some thoughts on. > > If you are using GMSK over a wireless channel, and if you can choose only > one feature to improve your system sensitivety, will you choose adaptive > equalizer or channel coding? Which provides more benefit? > > GSM has both while DECT seems to have none.
It depends on the channel response but lets pretend that it is AWGN. GMSK has controlled levels of ISI introduced to constrain the bandwidth of the signal. The BT of the Gaussian LPF is what determines the ISI. For values of BT less than 0.33 or so, the ISI has to be removed with an equalizer of some kind (decision feedback, MLSE). For BT of 0.5, the ISI can probably be ignored. John
> >mguo wrote: >> This is a very nice forum and a lot of knowledgeable people here. I
have a
>> question don't know if any experts can share some thoughts on. >> >> If you are using GMSK over a wireless channel, and if you can choose
only
>> one feature to improve your system sensitivety, will you choose
adaptive
>> equalizer or channel coding? Which provides more benefit? >> >> GSM has both while DECT seems to have none. > >It depends on the channel response but lets pretend that it is AWGN. >GMSK has controlled levels of ISI introduced to constrain the bandwidth >of the signal. The BT of the Gaussian LPF is what determines the ISI. >For values of BT less than 0.33 or so, the ISI has to be removed with >an equalizer of some kind (decision feedback, MLSE). For BT of 0.5, >the ISI can probably be ignored. > >John > >
Thanks for all the reply. Very helpful.
Since, it is a wireless channels, I suppose it depends on the delay
spread. For dense multipath channels, equalizer is a must to remove
ISI. But u need lot of overhead as u need to estimate the channel
adaptively.
Using channel coding , u may suffer in throughput. 
-SaiRamesh.