Font Size: a A A

Study On Efficient Data Gathering Based On Clustering For Wireless Sensor Networks

Posted on:2019-09-16Degree:DoctorType:Dissertation
Country:ChinaCandidate:S C SuFull Text:PDF
GTID:1368330596451703Subject:Control theory and control engineering
Abstract/Summary:PDF Full Text Request
The Internet of Things(IOT)is the third revolution of information technology following computers and the Internet.It is highly valued by governments,enterprises,and academia around the world.As the important carrier of data perception in IOT,Wireless Sensor Network(WSNs)is the basis for the application of IOT,and has shown great value in various fields of society.The main task of WSNs is to gather the sensed data from the external environment,involving information perception,wireless communication,data processing and other technologies.Due to the limited computing power and energy resources of nodes,how to use limited resources of nodes to collect data high-efficiently is of great importance.Combining the clustering and the mobile sink technology,this dissertation mainly focus on the data collection in static WSNs and in WSNs with mobile Sink,to explore some efficient data collection approaches based on clustering.The research works of this dissertation are as follows:(1)In order to balance energy consumption and prolong the network lifetime,the cluster formation process of nodes is modeled as a fuzzy partition of sample space in this paper,and an optimal clustering algorithm based on fuzzy C-means(FCM)is proposed.Firstly,the traditional FCM algorithm is improved from several aspects:the high-density region is obtained from the density parameter of nodes,which the initial clustering center is selected from;the optimal cluster number is obtained by the posterior fuzzy pseudo-F statistic method;the constraint conditions of membership value are broadened.On this basis,the improved FCM algorithm is then used for optimal clustering of WSNs.In the new round,the structure of every cluster remains unchanged,and the cluster heads are reselected based on the goal function of energy balance.In the data transmission phase,the member nodes of the cluster send the sensed data to the cluster head by a single-hop manner,the cluster head then combines the received data,and transfers the data to the base station according to the selected transmission cost function.The results of simulation experiments show that the algorithm can effectively balance the energy consumption of the network,improve the energy hole,and prolong the network lifetime while reducing the energy consumption of the nodes.(2)An energy-efficient data gathering algorithm based on unequal clustering is proposed for large-scale WSNs.Firstly,a hybrid hierarchy cooperation algorithm of genetic algorithm and particle swarm optimization(H~2GA-PSO)is studied to deal with large-scale optimization problems.Then,the H~2GA-PSO algorithm is applied to the clustering of WSNs,and the adaptive function is designed with the goal of energy balance.Consequently,a novel data collection algorithm based on non-uniform clustering is proposed.At the clustering stage,H~2GA-PSO is used for the optimal selection of cluster heads.The GA of basic level is used for global search to ensure global convergence.The upper level adopts the PSO algorithm with initial speed optimization to perform accurate local search so as to accelerate the convergence speed of the algorithm.At the routing stage for data collection,the member nodes send sensed data to the cluster head by a single-hop manner;after fusing the received data,the cluster head transmits the fused data to the base station according to the next hop object selected by depending on the transmission cost function.The simulation results show that the proposed method is superior to other methods in reducing and balancing network energy consumption.(3)A novel virtual force-based data aggregation mechanism(VFDA)with mobile sink is proposed.Firstly,the network area is divided into several grids,and the cluster is obtained based on the grid.The cluster head is elected according to the evaluation function for cluster head selection,and the cluster heads are responsible for data aggregation,forwarding,and cluster management.Then,the virtual force theory is used to analyze the virtual forces from the boundary,obstacles,empty areas,and the nodes including the cluster head,and the virtual resultant force of the mobile sink is so calculated.On this basis,the residence time for mobile sink and the coordinate of the next rendezvous point can be calculated based on the size and direction of force.As a result,the optimal moving path of the mobile sink is obtained.At each rendezvous point,the data aggregation tree is established with the sink used as the root node to collect the sensed data within the communication range.The simulation results show that the proposed method is effective even if for the WSNs with empty area or obstacle,and can obtain better performance than the existing algorithms in the aspects of efficient data aggregation,energy saving,and the path length of mobile sink.(4)In order to further balance energy consumption of the network and improve the efficiency of data collection,an efficient clustered data collection method(EBCDG)is proposed based on virtual grid and mobile sink.Firstly,considering the influence of the virtual grid boundary on the energy consumption of the nodes,a new cluster head evaluation model is established.In this model,the problem of cluster head selection is regarded as the problem of multi-attribute decision-making based on relative entropy,and the cluster head is optimized.Then,an optimal path selection strategy is proposed,and the allocation mechanism of residence time for mobile sink is adopted to improve energy consumption and data transmission delay.Finally,the EBCGD method is used to compare with other methods proposed in previous chapters of this paper.The experimental results verify the efficiency of EBCDG in data collection,and also comprehensively analyze the performance of other methods for data collection proposed in the dissertation.Finally,the research work of this dissertation is summarized,and the future work is prospected.
Keywords/Search Tags:Wireless Sensor Networks, data collection, clustering routing, mobile Sink, fuzzy c means, Genetic Algorithm, Particle Swarm Optimization, virtual force, energy efficiency, virtual grid
PDF Full Text Request
Related items