Fusion method of radar data and IFF data based on NMF

A new fusion method of radar data and IFF data based on nonnegative matrix factorization (NMF) is proposed in this paper, due to its strong part-based representation capability. The identification data from each sensor are put into one column of the input matrix and the fusion is realized by the converging process of a cost function. The proposed fusion method does not only show better performance in the credential level, but also has the advantage of limited computational load. The proposed method is simulated to validate the effectiveness and significant improvement in comparison with the data fusion algorithm under the framework of D-S evidential theory.