Question for fractionally spaced equalizer

Hi

I have a question for fractionally spaced equalizer

what is different fractionally spaced equalizer vs. symbol spaced equalizer

fractionally spaced equalizer(FSE) is oversampling

how to FSE is operating??


Do you have any reference???

To me it sounds like;

You receive a signal and you down-convert it to baseband again (I mean you get rid of the carrier). Then you have to apply a sampling process to convert it back to digital (1s and 0s). One option is to sample it at every symbol period (i.e. symbol spaced equalization). Well, it makes sense as long as you send one symbol at every symbol period However what if the signal is distorted due to very severe channel conditions? Then sampling once at each symbol period might mislead you. So to overcome such a problem, take, lets say, 10 samples in one symbol period. If the 1 or 2 of them are misleading, dont worry, you have 8 correct results to make the right guess upon which bit is transmitted. That is fractionally spaced equalization and its better but hard to implement than symbol spaced.
Sorry I dont have any technical reference for that, what I have is this story.