Hardware implementation of carrier and timing recovery circuits for wireless receivers

On Wednesday, 10 April 2013 11:14:17 UTC+10, Vladimir Vassilevsky  wrote:
> On 4/9/2013 5:54 PM, Udesh wrote:
> 
> > On Wednesday, 10 April 2013 01:43:21 UTC+10, Vladimir Vassilevsky  wrote:
> 
> >> On 4/9/2013 6:15 AM, Udesh wrote:
> 
> >>
> 
> >>> On Monday, 8 April 2013 05:52:18 UTC+10, Vladimir Vassilevsky
> 
> >>> wrote:
> 
> >>
> 
> >>>> On 4/4/2013 11:32 PM, James wrote:
> 
> >>
> 
> >>>>> I'm looking at ASIC and CMOS implementations on carrier and
> 
> >>>>> timing recovery for wireless receivers.
> 
> >>
> 
> >>>> What receivers?
> 
> >>
> 
> >>> Mobile wireless receivers.
> 
> >>
> 
> >> There are several dozens of different standards and no less then zillion
> 
> >> of proprietary platforms. Which one?
> 
> >>
> 
> 
> 
> > I want to start from some where. Could you please consider GSM ?
> 
> 
> 
> GSM is designed for batch processing. That is, burst signal is sampled 
> 
> in a buffer and then processed. Sync and carrier are established by 
> 
> search for max. likelihood using predefined sequence in the middle of 
> 
> the burst.
> 
> 
> 
> Vladimir Vassilevsky
> 
> DSP and Mixed Signal Designs
> 
> www.abvolt.com

Could you please suggest some specific documents where the above mentioned algorithms are implemented ? Can I have your email address ?

Thanks.

uOn 4/9/2013 10:28 PM, Udesh wrote:
> On Wednesday, 10 April 2013 11:14:17 UTC+10, Vladimir Vassilevsky
> wrote:
>>
>> GSM is designed for batch processing. That is, burst signal is
>> sampled in a buffer and then processed. Sync and carrier are established
>> by search for max. likelihood using predefined sequence in the middle
>> of the burst.

Besides, there are two different situations: initial synchronization and 
maintaining sync while in operation.

>
> Could you please suggest some specific documents where the above
> mentioned algorithms are implemented ? Can I have your email address
> ?

Refer to TS GSM (available from ETSI).
However, nobody is going to open specific details of particular 
implementations.

Vladimir Vassilevsky
DSP and Mixed Signal Designs
www.abvolt.com

I'd say Gardner's book is good, but it's not going to help much with modern techniques, and it's hard to find something that really does.  You're not really going to implement anything like it's shown in a textbook.  You can use lookup tables, for one thing, to take the arctan of received samples, which you can average to find how fast the constellation is rotating.  And you don't really need to do traditional PLLs--with a DDS, you can change phase instantaneously, which you can't do with a textbook PLL.