Comment and further results on "Efficient frequency offset resolution for OFDM systems based on ML principle"

This comment points out four inaccurate formulae and one problem in performance comparison of different estimators. The correct formulae is provided and the extension to the odd integer frequency offset is indicated, our simulations have shown that the simulation results exhibit good agreement with the new formulae more than the original formulae in Jie Zhang et al. (2004). In addition, the appropriate subcarrier number is obtained by the new formulae using MIP (minimum incorrect probability) principle.