Low-complexity Subspace Tracking Based Channel Estimation Method for OFDM Systems In Time-Varying Channels

In this paper, a group of low-complexity subspace tracking based pilot-aided channel estimation methods for orthogonal frequency-division multiplexing (OFDM) systems is studied. These methods are based on a parametric channel model where the channel response is characterized as a collection of sparse propagation paths. Considering the slow variations of the channel correlation matrix's signal subspace, we first translate the estimation of channel parameters into an unconstrained minimization problem. Then, to solve this minimization problem, a novel Kalman-filter based subspace tracking method is proposed, which employs the constant signal subspace to construct state equation and measurement equation. In contrast, other two adaptive filters, LMS and RLS, are applied and evaluated. These three methods constitute a group of low-complexity subsapce tracking schemes. It is shown that the proposed Kalman filter method is able to effectively track the time-varying channels, and outperforms LMS and RLS methods with large Doppler frequency spread.