Localization Technology And Application Based On Elman Neural Networks Of Wireless Sensor Networks

Wireless sensor network (WSN) is known as one of the most influentialtechnologies in the twenty-first century. And in recent years, WSN is a very populartechnology in the research and application at home and abroad. Involves varioussubject field, WSN has very important application value in national military andeconomic construction. In many of the wireless sensor network applications, that makessense if we combine the monitoring information with the actual position. So the nodesself-positioning technology became the key technology and the hot research, which aresignificant to the WSN’s research and applications.However, the cost、the volume and the power consumption constraints puts forwarda new request and challenge to the research of the WSN’s location technology. How toreduce the volume、the cost and the power consumption of the nodes on the premise theposition precision become the main emphasis of the research.The paper focuses the subject on the localization of wireless sensor networks. It tellsthe WSN’s related knowledge. We can get intuitive grasp though the discussion of thearchitecture、 features and applications of the network. The article detailed thelocalization of wireless sensor networks. The basic calculating nodes location methodsinclude trilateration、triangulation and maximum likelihood method，which are the basisof all the location algorithm. The location algorithm can be divided into range-basedalgorithm and range-free algorithm according to whether based on ranging. Although theformer has a high positioning accuracy, it is not appropriate large-scale use for that itneed extra hardware. The latter does not need extra hardware, but most of the data comefrom the algorithm are estimation of locations. If the positioning accuracy is not verystrict, the range-free algorithm has major advantages.The Elman neural network based distance model for WSNs is the importantcomponent of the paper. It is successfully to apply the Elman neural network to thelocation research of the WSNs. Elman neural network use the supervised learningalgorithm. And it belongs to the dynamic recurrent neural network. It is characterized bythe especial relation unit, which is used to remember the data before form the network. And with that the training will be completed speedily. The data acquisition and thesimulation result show the Elman neural network model has very high computationalspeed and allocation accuracy; it will have a good application prospect.Another important point of the paper is presents a self-localization system forwireless sensor networks used in logistics warehouses, which can be used to the workteam and vehicle. And the paper actual designed all kinds of nodes and their function.The sensor nodes use the rechargeable Li-ion batteries. So that is no need to commute thenodes will dead for energy is used up. The application of WSNs in logistics warehousessolved the problem that the GSP can’t be used indoors. Not only position but also cankeep an eye on the warehouse’s environment，the WSNs will certainly be a weapons tologistic warehouse management.