Almost blind channel estimation using hidden training

In this paper, a new method to perform channel estimation is presented. It is shown that accurate estimation can be obtained when a training sequence is actually arithmetically added to the information data as opposed to being placed in a separate empty time slot - hence the words 'hidden' and 'almost blind'. A closed form solution for the channel estimation variance is derived. A procedure is given to obtain training sequences that result in channel estimation independence of both the channel characteristics and modulation format. The problems of blind synchronization and dc offset are solved. Finally, from the simulations performed, the new algorithm is very competitive with those using traditional training, and outperforms all that are totally blind.