A method of individual radio transmitter identification based on collaborative representation

To make use of the correlation among the samples, a method of individual radio transmitter identification based on collaborative representation is put forward in this paper. Firstly, we extract the square integral bispectra (SIB) features to characterise the radio transmitters in the feature space. Secondly, based on collaborative representation, we compute the sparse coefficients with low computational complexity. At last, benefiting from the discrimination information of coefficient, the classifier is constructed for the final individual radio transmitter identification. On the actual collected FM radio dataset, the experimental results verify the effectiveness and robustness of our method.