A specular approach to MIMO frequency-selective channel tracking and prediction

We propose a specular approach to MIMO channel tracking that yields a parsimonious channel representation where the time and space components of the channel variations are separated by singular value decomposition (SVD).We show that this separation en- ables simple short-term channel prediction. This is es- pecially useful for systems using channel state informa- tion at the transmitter, since the latency necessary to the feedback process can be mitigated. We validate our approach through the use of computer simulations.