A novel signal recognition algorithm based on SVM in cognitive networks

In cognitive networks, cognitive users can access to the frequency spectrum, which may lead to interference on the primary users. These so called hostile signals are from hostile users, which spectrum sensing is not able to identify. These hostile users use spectrum at their own without considering communication rules will bring unnecessary interference to cognitive networks. In this paper, we propose a novel signal recognition algorithm based on support vector machine (SVM) in cognitive networks. Through extracting the signals' features we analyze and identify hostile signals. The proposed algorithm has two steps: 1) Main recognition achieves the classification of most signals; 2) For the rest of signals, secondary recognition will use a new characteristic to recognize again. Simulation results show that the signals can be successfully recognized in low SNR environment.