Channel estimation and tracking using implicit training

In this paper, a new method for channel estimation using superimposed (implicit) training is proposed. By adding the uncorrelated training to the information sequence at the transmitter, an accurate estimation can be achieved without loss of information rate. We prove the effectiveness of our method and a closed form solution for the estimation is derived. In addition, an adaptive algorithm is derived to track the time-and frequency-selective (doubly selective) channel. The performance of the method is then compared with existing approaches, and computer simulations show that the proposed algorithm exhibits good estimation behavior.