Research On Technologies Of Energy-efficient Clustering Data Gathering In Wireless Sensor Networks

With the development of wireless communication technology,embedded computer technology and MEMS technology,wireless sensor network has been widely used in people's lives.Data gathering is the basis of all kinds of wireless sensor network applications,and it is an important way to connect the physical world and the information world.At present,the technology of clustering data gathering has the advantages of simplicity,flexibility,energy efficiency,easy management and strong expansibility,and it has received extensive attention in academia and industrial world.In this dissertation,we focus on the technology of clustering data gathering for wireless sensor networks.According to the specific application,the suitable clustering data gathering algorithms are designed.They can not only improve energy efficiency of network,effectively prolong network lifetime,but also guarantee network QoS-aware performance requirements.As the research and application of wireless sensor network are closely related to each other,the network application scenarios and the performance requirements of data gathering are completely different in different applications.Based on the review and analysis of related works on clustering data gathering in WSN,this dissertation has focused on the following four aspects of research work,in which different network deployment scenarios(uniform node distribution or non-uniform node distribution)and the need to consider other QoS performances(network coverage ratio and reliability of data gathering)in addition to improving energy efficiency are discussed.Firstly,an unequal clustering data gathering algorithm based ring(UCDGAR)is proposed.Based on the premise that other QoS parameters are not required in uniform node distribution,this data gathering algorithm solves the problem of rebuilding the cluster topology periodically to result in more energy consumption.In UCDGAR,the optimal clustering parameters are theoretically computed when the network energy consumption is minimum.Furthermore,the network is divided into unequal cluster topologies by these optimal clustering parameters.Finally,once the cluster topology is established by the CH rotation algorithm,the role of CH is periodically changed.It means that the cluster topology is not rebuilt so that energy consumption is efficiently decreased and network lifetime is prolonged.Secondly,an unequal cluster data gathering algorithm with optimal cluster size(UCDGAOCS)is introduced.Based on the premise that other QoS parameters are not required in non-uniform node distribution,this data gathering algorithm solves the problem of the more message cost and the bad competition radius in the unequal clustering data gathering algorithms.Compared with the traditional unequal clustering data gathering algorithm,UCDGAOCS adopts time broadcast to substitute traditional discussion mechanism to select CH,which decreases the message cost to save the energy of network.Furthermore,when energy consumption of network is minimal,the optimal cluster radius is theoretically computed by which the clusters with optimal cluster are established to improve the balance of network energy consumption.Finally,during inter-cluster multi-hop routing,energy density factor is introduced,and the location technology of sensor nodes can implement directional routing discovery,which can not only effectively avoid the blind flooding of the routing probe packet,but also balance the energy consumption to prolong the network lifetime.Thirdly,a coverage-aware unequal clustering data gathering(CUCDGALS)is explored.Based on the premise that other QoS parameters are required in uniform node distribution,this data gathering algorithm solves the problem of the huge difference of energy consumption between CH and common nodes and ensuring target nodes coveraged for long time.At first,the coverage overlap factor is proposed as the main parameter of CH competition and relay nodes between clusters.The goal is that the nodes with large overlap factor have more possibility to become the headers and relay nodes.Furthermore,for alleviating the huge differences of energy consumption betweenCHs and common nodes,two CHs,the master CH and the slave CH,are selected by Ordered Weighted Averaging.The responsibility of master CHs is receiving data from their own clusters and other CHs,meanwhile,fusing these data,and that of slave CHs is transmitting data to next hop nodes.CUCDGALS can not only balance effectively the huge difference between the CH and the member nodes to prolong the network life,but also can improve the coverage ratio of nodes.Finally,an unequal clustering data gathering algorithm based on multiple criteria decision making(UCDGAMCD)is presented.Based on the premise that other QoS parameters are requred in non-uniform node distribution,this data gathering algorithm solves the problem of the imbalance of energy consumption and preventing node failures and packet loss.In UCDGAMCD,the multiple criteria decision making method based on intuitionistic fuzzy analytic hierarchy process and hierarchical fuzzy integral is used to synthetically evaluate multiple influence factors of CH selection Furthermore,a CH competition radius with dynamic adjustment is explored.In the scenario of non-uniform node distribution and heterogeneous energy,UCDGAMCD can establish the unequal clustering structure with energy efficiency and reliability.Moreover,in the process of transferring data between CHs,a routing algorithm with proportional allocation is proposed.UCDGAMCD can not only balance the energy ceonsumption and prolong network lifetime,but also improve the reliability of data gathering.