A novel method on pilot selection for sparse channel estimation in OFDM systems

Abstract

A novel pilot pattern selection method is proposed in wireless OFDM systems which can be used for sparse channel estimation. It is that cumulative coherence function is proposed to select optimal pilot patterns. This proposed method can characterize the measurement matrix which is made up of pilot patterns well, it can also remedy the drawbacks that better pilot patterns maybe ignored under the metric of coherence parameter. Simulation results demonstrate that optimal pilot patterns will be obtained by the novel method, and better channel estimation performance can be acquired.