Data Collection Method In WSSNs Based On Semi Supervised AP Clustering

Water Surface Sensor Networks(WSSNs)are wireless sensor networks composed of multiple sensor nodes that have specific tasks(such as detecting the physical properties of water surface)deployed on water surface.WSSNs can be widely used for water quality monitoring,oil production monitoring,water pollution control,which has important application value and development prospects.In WSSNs,the sensor nodes are usually deployed intensively only in the key monitoring area.In other areas only a small number of nodes are arranged.Generally,they have the characteristics of sparse deployment.At present,water surface mobile sink is widely used to complete the data collection,in which path planning for the mobile sink to collect the network data completely and efficiently is a fundamental key topic.In response to this problem,a WSSNs data collection method based on semi supervised AP clustering is proposed in this paper.Theoretical analysis and extensive experiments demonstrate that the proposed method can make excellent performance in the data collection of WSSNs.The main problem studied and major innovations in this paper are summarized as follows:(1)A data collection method in WSSNs based on semi supervised AP clustering is proposed in the paper.The method firstly constructs the similarity matrix based on the Euclidean distance between nodes and the residual energy of the nodes.Then the Responsibility matrix and the Availability matrix are iterated and updated constantly according to the similarity matrix,so that WSSNs can be divided into multiple clusters and the representative node in each cluster is selected.And the preferences are modified by dichotomy to adjust the clustering until the results satisfy the set conditions.Finally,the final path is formed for all representative nodes of clustering.(2)Extensive simulation experiments are designed using the above method on different network scales and compared with the existing representative methods to verify the advancement and feasibility of the method.According to the comparison results,the method proposed in this paper can effectively collect the data of WSSNs with the advantages of fast data collection speed,high energy-distance efficiency and long network lifetime.The method proposed in this paper for the WSSNs data collection has been proved to be effective by theory and experiment.It has some practical value for engineering application in this field.