Font Size: a A A

Research On Key Technologies Of Data Collection And Localizaiton In Wireless Sensor Networks

Posted on:2010-06-07Degree:DoctorType:Dissertation
Country:ChinaCandidate:W S TangFull Text:PDF
GTID:1118360305982691Subject:Computer Science and Technology
Abstract/Summary:PDF Full Text Request
In recent years, Wireless Sensor Networks (WSNs) have been one of the hot study spots in IT field. It was regarded as one of 21 kinds of most important technology and one of top 10 kinds of changing-world technology in 21 century, which was reported by BusinessWeek and MIT Review in USA. Further, unique characteristics of WSNs, such as large-scale, dynamic, self-organized, good reliability, application related, and data-centric, make it widely used in military, environment monitoring, forecast system, medical treatment and nursing, intelligent home-application, structure status monitoring, etc. Presently, there were lots of studies in Sensor Network Communication Protocol, its support and application technology, and gained many achievements.Two issues were studied in this thesis, one is the fairness control of data congestion in WSN and the strategy of maximizing network lifetime, second is the precision issue of RSSI fingerprints database localization technology and mobile node self-tracking.First, for the issue of fairness congestion control in WSNs, an Algorithm called EFFRC was proposed to realize the effective data transmission. The basis idea is: based on event-entric for WSN, allocate bandwidth according to the calculation of node's event information amount, so as to avoid the problem of redundancy data transmission by traditional method. In the calculation, being the fact that the node's event information amount is proportional to its potential area, to facilitate calculating the area of the event region, it is subdivided into several quadrilateral ones, and then the area of the region is the sum of the area of all separate ones.Second, for the problem of energy consumption of Heterogeneous clustering data collection network which regards maximum lifetime as target, proposed a finer energy consumption model and the new calculation method based on this new model for single-hop and multi-hops transmission scale, the network lifetime is prolonged accordingly. The basis idea is: further analyze the unbalanced status for nodes energy consumption during data transmission, then present the finer model based on the circle energy consumption model; based on this new model, we can optimize the parameters of nodes deployment strategy. After that a transmission algorithm mixed with single-hop and multi-hops prolonging the network lifetime is proposed.Third, to the precision issue of RSSI fingerprints database localization technology, an algorithm called CBFIA based on CPN (Counterpropagation Networks) neural network is put forward to improve the location precision. The basis idea is: improve the traditional CPN based on space relationship of RSSI, set up relationship between RSSI vector and location vector, and to interpolate the fingerprints database. Node location can be calculated with the smaller granularity, so as to be more precise and more efficient.Finally, to the issue of mobile node self-tracking, we propose the DMMPLA to improve the precision and reduce the calculation complexity. The basis idea is: due to mobile node location were affected by many environment factors, the error will be very big when we use the movement state equation to forecast the mobile node location; In order to reduce the error, we can adjust the location information by wireless distance measurement, convert the mobile node location problem to state estimate issue, then get the calculation. When solve the equation, we set noise and state estimator as two sides of game theory, by calculating their saddle point to get the equation solution, so as to achieve the location of mobile node.
Keywords/Search Tags:Wireless Sensor Networks, Data Collection, Fairness Control, Network Lifetime, Energy Balance, Node Localization, Mobile Node Self-Tracking
PDF Full Text Request
Related items