Fourth-order Cumulant Based CSPRIT Algorithm

Shen Yi,Chao Hou
2007-01-01
Abstract:By analyzing the problems of CSPRIT algorithm based on second-order statistics under space correlated Gaussian noise,a new algorithm combining fourth-order cumulant and CSPRIT algorithm is proposed to process one-dimensional signal sources,such as the binary phase shift keying(BPSK) and M-ary amplitude shift keying(MASK).DOA estimation and beamforming construction can be achieved simultaneously in this way.Compared to the original CSPRIT algorithm,this improved algorithm based on fourth-order cumulant can depress space correlated Gaussian noise well and get more accurate estimation of signals.The validity of this algorithm is also proved by computer simulation.
What problem does this paper attempt to address?