A dynamic channel estimation in MIMO communication system

Channel estimation is essential in digital communication. In MIMO transmission it is augmented by the need of transceiver with an equalizing sequence in each symbol. And the non-frequency and stationary fading channel is a prerequisite for MIMO transmission in state-of -art, that´s the why MIMO technique is mainly used as indoor communication. In this paper we propose to use an adaptive multiple dimensional support vector regressor (AM-SVR) for estimating MIMO communication channels because of the good generalization properties the SVR presents for short training sequences. Instead of using quadratic programming (QP) procedures for solving the M-SVR, as is usually done, we present an iterative re-weighted least squares procedure for attaining the M-SVR solution, because it is a procedure simple to implement and allows a recursive formulation that is impossible for QP schemes. We analyze the performance of the proposed algorithm using computer simulations in mobile MIMO communication scenarios.