Research On Data Collection Method Based On Compressed Network Coding In Wireless Sensor Network

Today, Wireless Sensor Network(WSN) plays an important role in society production and scientific research. It integrates sensor technology, microelectronics technology and wireless technology, provides an effective tool to know the world and change the world for human beings. In WSN, each node collects local data, and after processing the data by some way, the node sends the data to convergence node by some routing path, then convergence node decodes the data by some algorithm. It’s different routing path selection method and different data-processing method that formed different kinds of data collection scheme. There are two kinds of data collection scheme in WSN: based on the specific route or based on network coding.The former is more reliable, which can sense data incrementally, but it makes energy consumption of each node in a disequilibrium state, which decreases the network lifetime; The latter can improve the balance of energy consumption of wireless network, but there is a "All or Nothing"(AON) problem in the decoder, which decreases the reliability of the network.This paper firstly introduces the structure of and the lack of WSN and typical data collection scheme; then illustrates the mathematical model of Network Coding and Compressive Sensing, including Random Linear Network Coding(RLNC) and noncoherent measurement and reconstruction algorithm of Compressive Sensing. By analyzing the internal relation between them and using the correlation of WSN sensory data of all nodes and the radio characteristics of wireless transmission, we finally set up a data acquisition scheme of WSN with high energy-effectiveness,named Compressive Network Coding(CNC).On the basis of CNC, the paper analyzes how convergence node reconstructs the sensory data. The results show that the data reconstruction of convergence node is gradual, and the success rate of CNC is higher than the general network coding scheme by more than 15% in reconstruction, which solves the problem of AON perfectly. Then by analysis of the network energy consumption of each node, it shows that CNC can greatly improve the balance of energy consumption of the WSN,thereby reduces the total energy consumption and improves the lifetime of WSN.Beginning with the same energy in each node, the lifetime of CNC is more than 2.5times of Collection Tree Protocol(CTP) in WSN.