| Wireless sensor network (WSN) has become one of the focuses of the IT fieldfor it’s broad application prospects. WSN connect the objective existence of physicalinformation to sensor networks and speed up the way for people to obtaininformation. WSN are usually composed by the base station and a number of sensornodes, sensor nodes are responsible for collecting the data, and pass it to the basestation according to the data collection protocol, the base station analyzes the dataand reports to the user.This paper analyzes the objective needs of the current WSN data collection andthe constraint conditions of sensor nodes, so proposes a data collection method witha mobile base station and energy-saving tree for WSN (DMMC) and an algorithmwhich can be applied in the data collection methods for cluster adaptive multi-sensordata fusion (DFDE).Firstly, this paper describes the concept of the WSN, the characteristics of thearchitecture and the practical application of the scene, after that, the concept of datacollection methods, classification and performance evaluation of the data collectionmethods is also mentioned, and also tells the information of the multi-sensor datafusion for WSN, for example, the concept of data fusion and effects, and the currentmulti-sensor data fusion algorithm categories.Secondly, aiming at energy-saving, energy-balance, and large-scale applications,a data collection method called DMMC is proposed by a mobile base station andenergy-saving tree for WSN data collection in the monitoring area. According to theenergy and location information, nodes in the area are divided into a great many ofclusters, each cluster elect a cluster head, cluster heads consist of a data aggregationtree which rooted by the base station, and data are transmitted to the base station bythe aggregation tree. For balancing the nodes’ energy, after one round collection, thebase station run to the next edge location of the area, and then the next round of datacollection is began. In addition, DMMC also optimizes the clustering from the energy-saving in the role of multi-hop transmission.Again, in the basis of clustering by DMMC methods of data collection, anadaptive data fusion algorithm is also proposed called DFDE, which based ondensity estimation and used for inner-cluster. The algorithm can be used for theunknown prior knowledge of inner-cluster multi-sensor data fusion of large amountsof data values to the target value measured by sensors, then using the quantizeddensity value and support density value as the weighting coefficient, and ultimatelygive the multi-sensor fusion formula.Finally, this paper also gives the simulation test for the proposed DMMC andDFDE, obtain simulation results and performance analysis. All of these have greatlyverified the effectiveness of them. |
