Data-dependent superimposed training based channel estimation and symbol detection using Zero Correlation Zone (ZCZ) sequences

We address the problem of channel estimation and symbol detection using data-dependent superimposed training (DDST) in the presence of dc-offset for both single-input/single-output (SISO) and multiple-input/multiple-output (MIMO) systems using a set of Zero Correlation Zone (ZCZ) sequences. The use of ZCZ sequences in the proposed method not only results in dc-offset free channel estimation but also ensures that the equalisation output is completely independent of the dc-offset. The proposed approach is shown to perform significantly better than existing methods for any value of dc-offset and hence removes the computational burden of having to estimate the unknown dc-offset.