Research And Application On Data Gathering Algorithm Based On Energy Limited Mobile Sink For Wireless Sensor Network

Wireless Sensor Network(WSNs)has been used widely.In traditional WSNs,the sensor node location is usually fixed,so those nodes around base stations will take more communication-load,which leads to fast exhaustion of energy called energy holes.To address this problem,mobile-node techniques are recently introduced by using mobile-node as gathering-node to collect data,where the equipments may be mobile tramcars,cars runing on bridge or AUV.However,mobile nodes actually have their intrinsic limitations on spatial and load capacity.Therefore,under these conditions,how to design robust networks for resource sharing is a very chanllenging and key problem to be solved,and studing it will bring many potential pratical applications.In this thesis,we do a series of studies on data collecting in WSNs under the specific meteorological scene,and further propose some optimization algorithms to improve the network performance as well as prolong the whole network lifetime.First,an interval multi-hop routing method is proposed to collect meteorological data on the both sides of the road.Under the scenarios on the road,WSNs nodes are randomly equiped on the both sides,where base stations are located in the starting point of the road and thus the "energy hole" problem is particularly prominent in this banding WSNs.To this end,this thesis develops LEACH-Interval Multi-Hop(LEACH-IMH)routing algorithm based on LEACH protocol.First,consider many factors such as the optimal cluster head numbers,node residual energy and the communication distance with base station optimize the cluster head election algorithm in the selection model.Second,use multiple hops intervals to transmit fusion data to reduce energy consumption of cluster heads around the base station.Experiments demonstrate that the proposed LEACH-IMH can effectively prolong the lifetime of network.Second,a mixed integer programming model is proposed to overcome the limits of mobile distance of sink.Mobile sink can solve the "energy hole" problem,but its energy is not infinite.Especially,when mobile distance of sink is limited,the moving path of the sink and the retention period at each node decide the whole network lifetime.This thesis attempts to model node energy,sink moving distance,communication loss and others questions as a mixed integer linear progragraming,and further transform the best mobile path of sink into an optimization solution.Experiments indicate that the proposed model is suitable for small or medium-sized WSNs with sustainable running time and increasing time cost for larger networks.Third,a distributed discretization algorithm of calculating the optimal sink path is proposed to improve the efficiency of large WSNs.For large WSNs,the above mixed integer linear programming takes a long computation time.To address this problem,this thesis proposes a distributioned discretization optimization strategy.First,distributively collect data for each node by considering the number of its neighbor nodes and the distance to the sink,which not requires the global routing information.Second,construct a new transform matrix by formulating the longest stay time problem under the limit of mobile distance into the solution of shortest path under constraints.The experimental results show that the proposed method significantly reduces the computational time while achieving a competitive performance with the above mixed integer linear programming algorithm in the task of large WSNs.Finally,the aforementioned models are systematized and then succesffuly applied into a real-world case of the Nuist campus.Meteorological WSNs are deployed at the Nuist campus to observe regional climate,test the network with multiple nodes,and validate the function of data collection including multi-hop transmission,instructions,etc.The real experiments indicate that this system can robustly solve the network optimization in the energy limit of mobile sink,and further meet the requirements of various service applications of meteorological WSNs.In conclusion,this thesis develops a series of studies on data collection in WSNs,and proposes some corresponding solutions especially for the case of the distance constraint on mobile nodes.The extensive experiments verify that the interval multi-hop routing algorithm and the distributed discrimization optimization algomizaation can balance the energy load of each node and prolong the network lifetime.More applications of the proposed models will be further exploered in future.