A New Spectrum Map Fusing Method Based on Difference Group Sparsity

Fusing spectrum maps from multiple radiation sensors can suppress undesired sporadic data caused by single-sensor sampling, and thus, the accuracy of spectrum map reconstruction can be improved. However, most current spectrum map fusing methods do not consider the electromagnetic signal propagation rules, so their performance enhancement is limited. In this paper, a new spectrum map fusing method based on difference group sparsity is presented, which can fuse spectrum maps reconstructed by multiple sensors. We first construct a spectrum map fusing model based on the sparsity of spectrum data. Then, according to the signal spatial path-loss model, we further analyze the group sparsity of the discrete spectrum map data, and design a fusion regularization term to suppress the outliers introduced by the single-sensor spectrum map reconstruction. Simulation results show that the proposed method can effectively improve the accuracy of spectrum map reconstruction compared with the state of the art.