基于特征融合的智能家居射频指纹识别技术研究

s the security of smart home wireless communication becomes increasingly tense, the existing recognition technology faces challenges such as insufficient characterization of device features due to complex multipath propagation, and high false alarm rate triggered by low intra-class recognition rate, etc. In this paper, we propose a deep feature fusion closed-set RF fingerprint recognition model based on feature fusion and attention mechanism. By fusing 13 kinds of expert features with 1DCNN features and combining the SE attention mechanism for feature weight allocation, the recognition rate of smart home devices is improved. Experiments on public datasets show that the method proposed in this paper achieves a recognition rate of 99.84%, which effectively solves the problem of insufficient device recognition rate in smart home environments.