Channel estimation using time-frequency techniques

We propose the use of time-varying signals and time-frequency techniques to estimate fast fading channels with unknown state information. Specifically, we apply the reassigned spectrogram to estimate multiple sinusoids resulting from transmitting and processing two linear frequency-modulated (FM) chirp pilots. The estimation procedure is complicated by the need to pair sinusoids resulting from the same multipath. This problem is overcome using a single nonlinear FM chirp pilot with power characteristics. The channel estimation simplifies to estimating linear FM chirps which we perform using a modified matching pursuit decomposition algorithm. Simulations demonstrate the increased estimation performance when pilot signals with nonlinear structures are used instead of linear structured ones.