Wireless Feature Extraction System for Cloud-Dust Based Intelligent System by Using Embedded Wavelet Packet Method

摘要

In the paper,a wireless feature extraction system for cloud-dust based intelligent system using embedded wavelet packet method is developed.In order to decrease the massive transmission data form receivers and transmitters of wireless sensors network,we reduce the data dimensions of sensors using embedded feature extraction method.The massive transmission data will be reduced via the process of wavelet packet decomposition and reconstruction.So it decreases the transmission loading of general wireless sensors network.Considering computing time of feature extraction,the circuit hardware of wavelet packet is designed in the paper.It effectively reduces the computing time of feature extraction.Therefore,the wireless feature extraction system using embedded wavelet packet method makes the transmission of wireless sensors network work effectively.At last,we prove the data reduction of transmission,the shorter computing time of feature extraction and the computing accuracy of wavelet packet hardware by the experiments.It can greatly improve wireless sensors network with large number of nodes more efficient transmission,and the sampling rate of sensors nodes,without causing the transmission delay and data errors.So it will make wireless sensors network more applications with higher precision,higher computing speed,lower power and lower cost in the future.