Modified Orthogonal Matching Pursuit for MISO Channel Estimation in DVB-T2

Channel estimation in OFDM-based systems, such as DVB-T2, can be performed with compressed sensing techniques since the channel frequency response can be considered a sparse signal. In this paper we present a novel modification to the orthogonal matching pursuit algorithm which avoids illconditioned (partial) DFT matrices and hence mitigates noise enhancements. Simulations for MISO in DVB-T2 show that it can offer substantial performance gains at a modicum of increased complexity.