Skip to Main Content
Because of its multi-resolution analysis performance and great flexibility, wavelet packet transform is suitable for the narrowband interference. It can be realized by quadrate mirror filters, which decomposes the received signal into the n level to obtain a complete binary tree. According to received signal characteristic, we select an appropriate threshold. Compared with the threshold, the value including NBI is set to zero, and calculate the cost value of each node by employing the cost function. Each node is marked with its cost value. In calculating best basis process, these “marked” are modified from bottom toward top, we select a best basis set finally. Using the set and the position in the tree, the NBI free signal is reconstructed. The simulation results show the algorithm further improves the narrowband interference suppression performance of the DSSS system.