If somebody has good knowledge about the synchronisation method (related to the OFDM topic) developed by Schmidl and Cox, then you will probably be able to help me with my question about their synchronisation method.
I understand that their method does apply to OFDM and wireless LAN. And there are articles discussing their synchronisation technique, which involves the transmitter sending TWO identical sequences in succession. And their technique is based briefly as follows:
At the receiving side, while the transmitted sequence will likely be altered by effects of the transmission channel - the main idea is that the received time-domain signal will exhibit two identical repeating sequences (in a row...ie. in succession). These are expected to show up every once in a while.
The follow-on idea is to allow the receiving side to use a correlation method to find (detect) the occurrences of the consecutive repeating sequence pairs ----- which (if they are found) can provide a method of synchronising with the incoming signal.
However, in OFDM wireless communications, the transmitted sequence typically involves up-conversion, ie. mixing values with a transmitter carrier signal. And at the receiver, down-conversion must be done. Any differences between the carrier frequency of the transmitter and the receiver will result in a carrier offset frequency (CFO).
A lot of articles explain the usage of the Schmidl and Cox method for the case of ZERO carrier frequency offset.... or no carrier frequency offset.
I'm thinking that the Schmidl and Cox method is not going to work when carrier frequency offset is non-zero, right? I was thinking that their synchronisation method relies on the receiver being able to find pairs of consecutive repeated sequences (ie. linked to the Schmidl and Cox pre-amble). This would certainly work nicely when there is no carrier frequency offset.
But, in the presence of CFO, the down-converted received signal will still be modulated to some extent (like AM suppressed carrier modulation). This is all due to CFO. For this case, the modulation due to CFO spoils the party, right?
That is - we would ideally like to see repeated consecutive sequence patterns in the incoming received (down-converted) signal in the time-domain. But a received sequence affected by non-zero CFO is not going to show the observable or measureable pairs of repeating sequences ---- because CFO introduces instantaneous time-dependent amplitude variations in the received signal - right?
If CFO does spoil things, then does anybody know if the Schmidl and Cox method is still workable with non-zero CFO? That is ----- what needs to be done or considered if we are to apply this method in practice?
Thanks for your help and comments in advance!
