The autocorrelation matrix eigenvalue of the signal reflects the power of each direction,and different characteristic values correspond to different signal spaces.With interference present,large eigenvalues generally corresponds to interference signal subspace,and small eigenvalues to signal subspace.Removing the corresponding interference subspace can remove strong interference signal.Based on this,this paper puts forward an anti-interference algorithm using the matrix eigenvalue that can effectively eliminate obstructive and tone-like interferences.This algorithm has the advantages over common FFT trapping-based algorithms of smaller amount of computation and simple logical realization,thus decreasing the complexity of anti-interference algorithms.