Improved suppression algorithm of digital narrowband interference in FFT domain

An adaptive FFT method for suppressing digital narrowband interferences based on double threshold is proposed in DSSS communications. The proposed algorithm utilized the threshold parameter of consecutive mean exciser (CME) to set a high and low threshold. First, the lower threshold groups the adjacent samples above the lower threshold into the same group, called as a cluster. The largest element of a cluster is compared with the upper threshold. If the largest element exceeds the upper threshold, it is decided that the cluster corresponds to a NB signal, while if the largest sample is below the upper threshold, the cluster corresponds to the noise and possible desired signal. Theoretical analysis shows the importance of the double threshold, which avoids the BER performance degrades at low JSRs. Numerical simulations show the exactness of the theoretical analysis and the validity of the proposed method.