Interpolation of Sparely-Sampled Time-Variant Channel Response in 1D Trajectory Utilizing Ensemble Kalman Filter

Stored channel measurement can be utilized for developing site-specific radio channel emulator with high accuracy. The channel emulator is expected to synthesize temporal fading characteristic in terms of a time-variant channel impulse response (CIR) of the mobile station (MS) at any position in space. However, such continuous characteristics is hard to acquire directly from the measurement because only a set of discrete CIR samples could be sparely-captured. In this work, interpolation technique based on ensemble Kalman filter (EnKF) is proposed to reconstruct the continuous CIR profile given a sequence of undersampled measured CIRs. The proposed EnKF model was utilized to estimate and track fading parameters between two adjacent measured CIR. The interpolated CIR is then computed from those estimated parameters. Through experiment, performance of EnKF-based interpolation is evaluated in the simulation using 3GPP channel model. The advantage and limitation of the proposed interpolation scheme is discussed based on the simulation result.