Research On WSNs Environmental Monitoring Method Based On Improved Kriging Algorithm

With the development and progress of information technology,Internet of Things(IoT)has integrated into all aspects of social life gradually.It is widely used in industrial production,environmental monitoring,health care and many other fields.Wireless Sensor Networks(WSNs)are one of the infrastructures and key technologies in IoT.They have the characteristics of self-organization,large scale and dense distribution.And restricted by the cost of nodes,they have limited energy resources,computing resources and storage resources.Hence,how to improve network performance is the focus of application research in WSNs.This thesis mainly studies the environmental monitoring methods based on WSNs.Node scheduling is an effective way to extend the lifecycle of WSNs.However,node sleeping may cause the loss of monitoring data and the reduction of monitoring accuracy.Therefore,Kriging algorithm is introduced to estimate the missing data.The problem of ordinary Kriging algorithm is that the fitting accuracy of the semivariogram model is not high.So the NM simplex algorithm based on Gaussian variation is introduced in this thesis and an improved Kriging algorithm(IK)is proposed.Simulation results show that the improved algorithm is better than the ordinary one.Next,IK is applied to WSNs,and it is combined with an energy-balanced sleep scheduling algorithm.Then an interpolation estimation algorithm based on IK for WSNs is proposed.Simulation results show that,the proposed monitoring method can not only extend the network lifecycle,but also guarantee good monitoring accuracy.Finally,based on the Kriging interpolation estimation algorithm,a simulation system of WSNs Environmental monitoring is designed and implemented,and the demonstration is carried out.