Light Wiener filters for estimation of slowly varying wireless channels

Abstract

We derive reduced order (light) Wiener filters for slow flat fading channel estimation in pilot aided Alamouti receiver. The concept is based on the approximation of the channel auto-covariance function to zero and second order Taylor series for the cases when there is only partial knowledge about channel statistics at the receiver. Moreover this approach reduces computational effort in design of the estimation filter and its implementation. The theoretical Mean Square Error (MSE) is developed for each case, verified through simulation and compared to that of a full Wiener filter. Results are then applied to evaluate performance of Alamouti coding Multiple Input Single Output (MISO) transmission system with imperfect Channel State Information (CSI) at the receiver side and pilotbased estimation scheme.