FBMC/OQAM system channel estimation based on VSCBOMP algorithm

Abstract

Compared with conventional orthogonal frequency division multiplexing (OFDM) systems, filter bank multi-carrier/offset quadrature amplitude modulation (FBMC/OQAM) has the advantages of low bit error rate and high spectrum efficiency, but there are few studies on the use of compressive sensing in order to estimate FBMC/OQAM channels recently. A channel estimation algorithm named variable step-size conditional backtracking orthogonal matching pursuit (VSCBOMP) is proposed for FBMC/OQAM systems. This algorithm selects several atoms in each iteration and backtracks when the support set is full in order to get rid of atoms with poor correlation. The idea of variable step-size is added simultaneously to improve the operation efficiency. The algorithm is then simulated in sparse random signals recovery experiment and channel estimation in FBMC/OQAM systems. The simulation result proves that the algorithm has shorter running time and higher estimation accuracy than conventional orthogonal matching pursuit algorithms.