A novel superimposed periodic pilot scheme for finite-impulse response (FIR) channel estimation
the linear minimum mean square error estimator (LMMSEE)
is investigated.The advantage of the method is that it is simpler to implement without sacrificing the data rate as it needs no information on the channel statistics.Theoretical analysis and computer simulation show that the proposed method are found to exhibit better performance than that of the linear square (LS) method at lower SNR values
provided that the number of pilot tones is sufficiently greater than the channel impulse response’s length.