面向智能物联的双路径机器学习调制模式识别技术研究

分析了一种面向智能物联的双路径机器学习调制模式识别方法 .以元学习网络为基础，将模糊分类和精确分类相结合，在低信噪比环境下，能扩增识别信号类型，同时保持较高的识别精度，有效地降低了恶劣环境对信号带来的影响.用理论分析和仿真结果证明了该方法的有效性.

A dual path machine learning modulation pattern recognition method for intelligent Internet of Things

In this paper a dual path machine learning modulation pattern recognition method was analyzed for intelligent Internet of Things. Based on meta-learning network and the combination of fuzzy classification and precise classification, the types of identification signals were increased, and high identification accuracy was maintained in the low signal-to-noise ratio environment at the same time which could effectively reduced the impacts of harsh environment to the signal. The theoretical analysis and simulation results demonstrated the effectiveness of the proposed method.