Symbol recovery using correlation

[babyeric]

Hi,

I'm trying to understand symbol synchronization. I understand that by using correlation of the received samples with the upsample preample samples, we can find out where is the starting point of the symbols. However how can we use correlation to find the exact sampling point of the symbols.

This is what I'm doing:
Preample signal --> Upsample preample signal by 4 --> RRC filter --> noise --> RRC filter --> correlation with upsample preample signal ...

Till this stage, i'm able to find out the starting point of the symbols but unable to find out what is the exact sampling point. Theoretically, assume by symbol rate is 1, so the sampling point is 4. But sometime due to noise, the exact sampling point might shift so is there a way to find out using correlation??

Hope experts out there can help. Thanks a million.

Eric

 

[cyteng]

You need to do timing synchronization to find the exact sampling point.