Joint interpolation for LTE downlink channel estimation in very high-mobility environments with support vector machine regression

In this study, the estimation of fast-fading long term evolution (LTE) downlink channels in high-speed applications of LTE advanced is investigated by the authors. A robust channel estimation and interpolation algorithm is essential in order to adequately track the fast time-varying channel response. In this contribution, the multipath fast-fading channel is modelled as a discrete, tapped-delay and finite impulse response filter. Using support vector machine regression (SVR), they develop an extended algorithm to jointly estimate the complex-valued channel frequency response in time and frequency domains, in the presence of fading and non-linear noise from the transmission of known pilot symbols. Furthermore, the channel estimates at the known pilot symbols are interpolated to the unknown data symbols by using the non-linear SVR approach exploiting kernel features. This study integrates both channel estimation at pilot symbols and interpolation at data symbol into the complex SVR interpolation method. The bit error rate and mean square error performances of the authors' fast-fading channel estimation scheme is demonstrated via simulation for LTE downlink with 64-QAM modulation and 500 km/h velocity under non-linearities.