Time-interpolated channel estimation for increasing channel capacity of superimposed training-based precoding schemes

This study proposes a time-interpolated channel estimation method for the superimposed training-based precoding scheme in slowly time-varying channels. The regular inclusion of a superimposed training sequence, which is necessary for explicit time-varying channel estimation, reduces the power allocated to information transmission. Therefore time-interpolated channel estimation, where the transmission without a superimposed training symbol is performed during a few coherence times, makes very efficient use of transmitting power and possible gains in terms of capacity. A closed-form solution of the periodicity of time-interpolated channel estimation was derived using this method together with a Markov channel model. The simulation results confirmed that the use of the time-interpolated channel estimation method improves the performance of channel capacity and possibly provides energy savings for mobile terminals.