The Research Of Compressed Sampling At Sensor Network Nodes

With the development and application of the Internet of things and Sensor Networks, theamount of data transmission and data collection is increasing, and the existing technology isfacing challenges. The reduction of the amount of data transmission and data collection isneeded.The theory called compressed sensing proposed new possibilities for signal samplingmethods. The theory states that, using the sparsity of the signal, we can reduce the samplingfrequency. As the compressed sensing theory is increasingly perfect, much attention has beenpaid to its application in the field of sensor network also.Firstly, this paper introduces the sensor network and the background of compressedsensing. The problems existing in the sensor network are introduced and the meaning ofcompression perception in sensor network applications is explained. And the existing ways ofthe realization of compressed sensing are introduced. Secondly, a measurement matrix whichisSuitable for sensor network is present. And simulation is conducted for it. The simulationresults show that this measurement matrix can be used for compressed sensing. At last, thispaper presents a random sampling method for sensor network，which is based on the principlecalled compressed sensing. According to the characteristics of sensor network, we design therandom sampling method to acquire the useful information in signal. Experimental resultsshow that the system can realize compression sampling. And then we optimized the system.The simulation results show that the optimized system can reconstruct the original signal withhigh accuracy.