A Learning Muti-Sources Weight Algorithm of Electromagnetic spectrum Cooperative Cognition

With more and more smart devices have been used in many fields, there are many complex electromagnetic behaviors or dangerous signals through these devices. In this paper, a novel framework to learning multi-sources weight algorithm(LMSW) is presented to electromagnetic spectrum signal cooperative cognition to recognize these signals. These signals of the target from different sensors are fused by a learning weight matrix through the whole train network. Different signals from different received sensors are assigned different contributions to the cognition or some other mission. Features of the fused signals are extracted by deep learning network automatically. The experiment results show that the more accuracy is got through the LMSW algorithm than other methods.