Training Signal Design for Channel Estimation in Monostatic Multiantenna Backscatter Communication

We consider forward channel estimation at a multiantenna hybrid access point supporting monostatic backscatter communication. We design statistical and channel-adaptive training signals to leverage the performance of a conventional forward channel estimation method using backscattered signals. The statistical design optimizes the training signal for linear minimum mean square error estimation of the correlated backscatter channel through the majorization-minimization algorithm. The channel-adaptive design selects an antenna sending the next training symbol by estimating the channel power of each antenna using the backscattered training signals available currently. The channel-adaptive training signals improve the performance of the forward channel estimation with a reduced training overhead in the high SNR region while the statistical training signals improve the performance in the opposite case.