Search

Aiming at the low recognition rate when SNR is low in single node, this paper proposed a distributed cooperative method to recognize different modulation types based on wireless sensor networks and realized correct recognition of four classic modulations in low-SNR environment. It used several collaborated sensors in networks to improve the performance of the system and based on the principle of lowest sensor overhead. It designed cooperative method according to the relationships between the ability of the characteristics with the SNR and the calculation of characteristics, and each node extracted part of the characteristics. It combined a feature vector by the features which sent to the central node. Then it used support vector machine classifier to recognize the signals. Simulation results show that the recognition rate of the proposed method is higher than the single-node and the existed multi-nodes method, overhead of sensor is reduced, and has better adaptability to SNR environments. [ABSTRACT FROM AUTHOR]