Channel estimation for an OFDM-based MIMO system

We consider the problem of channel estimation for an orthogonal frequency-division multiplexing (OFDM)-based multiple-input multiple-output (MIMO) wireless local area network system. We show that an existing channel estimation approach, which is based on a sub-carrier level orthogonal training symbol design and the finite impulse response channel model, can suffer from severe performance degradations for realistic OFDM channels. As such, we propose a new channel estimation approach, which employs a modified version of this training symbol design and a polynomial fitting/interpolation method. Our simulation results show that our new channel estimation approach significantly outperforms the existing one for the realistic OFDM channels.