You are not logged in.
In the MIMO-OFDM PHY, there's a "rxScaling" factor used for post-equalization scaling. Why does it depend on the number of training symbols?
Offline
Because our channel estimator takes the running sum of estimates from consecutive training symbols- the average without dividing. After the equalizer was a convenient place to put the multiplier that corrects the resulting scaling. Since the PHY driver knows the number of training symbols at run-time, it can also update the scaling as needed.
Offline