A Vibration Communication Method for Smart Devices Based on Convolutional Neural Network and Transfer Learning

Abstract

The invention is a vibration communication method for smart devices based on convolutional neural network and transfer learning. The invention relates to the technical field of vibration communication of the Internet of Things. The invention performs bit stream group modulation and transmission of vibration signals on the sending end of the smart device; uses a beacon bit detection algorithm to determine the starting point of vibration on the receiving end of the smart device; and uses a three-axis accelerometer signal Perform principal component feature extraction to remove signal noise; perform convolutional neural network decoding on the acceleration signal after principal component analysis feature extraction to obtain the symbol label corresponding to the bit group; when the communication environment changes, perform transfer learning to improve the vibration signal recognition accuracy. The method of bit stream group coding combined with convolutional neural network decoding adopted in the present invention does not need to care about the influence of intra-symbol interference on transmission accuracy, and can speed up the vibration communication rate. Compared with the amplitude modulation and frequency modulation technologies, the present invention can be commonly used by various current commercial smart devices.

Description

Technical field [0001] The present invention relates to the field of Vibration Communication Technology of the Internet of Things, is a vibration communication method for intelligent devices based on convolutional neural networks and transfer learning. Background [0002] Smart IoT devices are booming, involving areas such as smart homes, smart cities and smart medical care, which have greatly changed and enriched people's lifestyles and effectively improved people's quality of life and service levels. Information interaction and communication are essential functions for smart devices to operate and use, and ensuring secure device communication is critical. At present, commonly used smart device communication methods include WiFi, Bluetooth, infrared, ZigBee, etc. They have their own advantages, but from the perspective of information security, they are vulnerable to wireless eavesdropping attacks. [0003] The proposal of vibration communication technology of intelligent equipme...

AI Technical Summary

Problems solved by technology

Therefore, as the bit rate increases, there is a possibility of vibration propagating between adjacent symbols and causing significant intersymbol interference

This limits the transfer rate of techniques based on switch control

In addition, vibration technology based on amplitude modulation and frequency modulation cannot be generally supported by current smartphones

Images