Channel and DC-offset estimation using data-dependant superimposed training

We address the problem of multipath channel and DC-offset estimation in communications systems. We extend the recently proposed data-dependent superimposed training (DDST) to include DC-offset estimation. Optimal design of the known superimposed sequence is also derived. The proposed pilot assisted method is compared to the conventional superimposed and time-division multiplexed training schemes. Consider a cyclic-prefixed single-carrier block transmission system operating over a frequency-selective channel. The latter is assumed to be time-invariant during each block but it may vary across the blocks. In order to avoid interblock interference, we assume the length of the cyclic prefix to be larger or equal to the length of the channel. At the receiver, after removing the cyclic prefix, the baud-sampled discrete-time baseband signal model for each received block (we omit the block index for notational).