Abstract:

In OFDM system,sparse channel estimation based on compressed sensing(CS)can make full use of the inherent sparse degree of the wireless channel,which can reduce the pilot overhead and improve the spectrum efficiency.Therefore,a new method based on hybrid genetic algorithm was investigated for the pilot design of CS channel estimation,which was based on the minimization of the matrix cross correlation in the CS theory.In this method,genetic algorithm was used to obtain the initial sub-optimal pilot sequence,and then combined with the pilot position and pilot power,each entry of pilot pattern could be sequentially updated and optimized to make the minimum correlation of measurement matrix.Simulation results show that the proposed method can ensure a better mean square error and bit error rate compared to the pseudo-random pilot design and the equal distance pilot design.

Key words: channelestimation, crosscorrelation, hybridgeneticalgorithm, measurementmatrix, compressedsensing,pilotpattern