| Popis: |
To measure the similarity between the fine features of communication transmitters, a method of individual communication transmitter identification based on maximum correntropy was put forward. Firstly, the square integral bispectra was extracted to characterise the communication transmitters, and then optimization function based on maximum correntropy criterion was constructed. Secondly, the half-quadratic technique transformed the nonlinear op-timization problem to a weighted linear least squares problem. Finally, for identifying the communication transmitters, the sparse coefficient computed by active set algorithm was utilized to construct the classifier. Using the FM radios with same manufacturer and model, the actually collected data sets verified the feasibility and effectiveness of the method. |
