Symbol-Blanking Superimposed Training for Orthogonal Frequency Division Multiplexing systems

While the single-carrier concept of superimposed training (ST) for channel estimation has already been applied to the very popular Orthogonal Frequency Division Multiplexing (OFDM), its single-carrier enhancement (Data dependent superimposed training (DDST)) cannot be so combined with OFDM. That is because the inherent structure of OFDM means that certain input symbols will now be permanently lost (or nulled) with DDST. So here for the first time, we propose a DDST-like approach that can now be applied to OFDM - Symbol-Blanking Superimposed Training (SBST). We show that by the simple expedient of combining (prior to transmission) these nulled symbols with symbols at other subcarriers and proposing two new receiver detection algorithms, the performance of SBST can very closely approach the traditional pilot tone assisted modulation (PTAM) of OFDM - but without the concomitant reduction in data rate associated with PTAM.