Research On Topology Control Of Wireless Sensor Networks Based On Improved LEACH Algorithm

Wireless sensor network(WSN),as a self-organizing network,it acquires regional information in real time through sensor nodes and can have the information processed.Sensor nodes have a certain capacity for collecting,transmitting and processing information.Sense nodes are usually in small size,with low cost but limited battery power,thus the number of them in network is large.In sensing area,nodes are randomly and evenly distributed.By analyzing the information nodes collect,WSN can achieve the goal of monitoring the regional environment.Sensor network has incomparable advantages of acquiring information in real time and we don't need to build infrastructure ahead.It has been so widely applied in various fields like medical treatment and environment.However,as it's not easy to replace the nodes' battery and also nodes' communication range is overlapped,it has been a hot topic on the research about how to reduce the redundant data in the network to improve the effective utilization of nodes' energy.Regarding how to reduce the energy loss of nodes and extend the life cycle of the network,this dissertation aims to overcome the drawbacks of traditional LEACH(Low Energy Adaptive Clustering Hierarchy)algorithm,and to bring up with an improved TLE-LEACH(The Least Energy LEACH)algorithm.This algorithm makes some improvements on deciding the optimal number of cluster heads in a network,selecting cluster heads and decreasing redundant data in a cluster.In addition,an idea about using intermediate node for data fusion and data forwarding is proposed to optimize the information transmitting path.The main work of this dissertation is listed as below:Based on traditional hierarchical network and the basis of the LEACH algorithm partitions,an optimal number k for clusters is calculated according to the energy loss in the current network.Redivide the network area based on cluster number k and select cluster heads based on energy optimization principle in the network,which can solve the cluster heads' distribution problem,either too intensive or scattered.Meanwhile,this dissertation introduces a concept of using primary and secondary cluster heads to prevent data loss from an accidental system down issue of the primary cluster head.The rotation of primary and secondary cluster heads reduces the unnecessary network energy loss caused by cluster head's election.Like traditional clustering,cluster-head nodes undertake the task of data forwarding and fusion.Energy consumption of the nodes depends on the volume of data communication between the nodes.Data fusion is an effective way to improve the credibility of the information and to reduce data redundancy.Each node in the network will build matrix w by calculating the trust from other nodes and giving different weights to the neighbour nodes for data fusion.In TLE-LEACH algorithm,through a single jump,nodes will transfer data to the cluster head nodes.The cluster head will firstly perform data fusion process and then send it to the base station.Transfer data by a single jump,energy loss is huge for node that is far from the cluster head nodes.So we need to optimize the transmitting path of information.Start from Floyd algorithm and Dijkstra algorithm,based on characteristics of nodes in WSN,find a way to optimize the path with special circle plus angle,to reduce the energy loss and extend the life cycle of the sensor network.Introducing intermediate nodes for data fusion.Intermediate nodes will process data fusion before transferring it to the main cluster head,which can reduce energy loss effectively for primary and secondary cluster head nodes on data fusion.It can also reduce the times of election for cluster head nodes indirectly,which has significant benefits to extend the life cycle of sensor network.Finally,to link with the actual project.Based on the research results above,a detection system of soil platform in harsh areas was designed and developed to visualize the information collected by sensors.The platform integrates the functions of human management,data detection management and timely monitoring of the soil.The system can display the collected environmental information of the detection area,and the system staff can improve the soil according to the data,which can be well applied in real life.