Research On The Data Storage Scheme And Applied Technology Of Wireless Sensor Networks

The fundamental task of Wireless Sensor Networks(WSNS) is to efficiently get valuable informations of the physical world and send them to the users. Wireless sensor network is data-centric, regardless of the application scenarios and the underlying hardware deployment of WSNs, users are most concerned about is efficiently transmite and store the data collected by the sensor nodes in order to facilitate post-processing. That is why data storage becomes an important research area in wireless sensor networks. The current data storage strategy is divided into three categories, such as local storage, external storage, data-centric storage. The main research area in this dissertation is data-centric storage, since it represents the development trend of the data storage technology.This dissertation analyzes the routing algorithms used in data centric storage. Geographic information routing protool is a classic routing protocol which has been widely used in data-centric storage schemes. This dissertation analyzes the geographic information routing protocol based on greedy forwarding and perimeter forwarding, from the perspective of topological analysis and energy efficient. The study pointed out that there is a large energy waste when the perimeter forwarding mode try to find the storage nodes, and then propose the approximate location algorithm which can reduce the perimeter forwarding. The basic idea of the algorithm is trim the perimeter forwarding process when the distance between the node and the destination location is close enough. The approximate location algorithm can effectively reduce the energy consumption caused by the perimeter walk, thereby makes the use of network energy more efficiently.In data-centric storage schemes, events may be stored in a small number of nodes, too much store or query operations to these nodes may make the energy or storage space consumed too quickly, resulting in storage hotspot issues. With regard to the hotspot issues, this dissertation proposes a virtual grid based multi-threshold algorithm dynamic allocate storage nodes can make the data of event stored on nodes more evenly. The basic idea of this algorithm is as follows:the event is hashed to a grid, and the grid dynamically allocate node to store the even according to the current storage space and energy of the node. It can prevent a node from being assigned too many storage tasks by setting the threshold and virtual coordinate, thus avoiding the hotspot issues accordingly. This algorithm makes the data distributed more evenly, thus improving the service quality of the network and prolonging the networks'live time.This dissertation also studies feasible energy waste in the data-centric storage. In wireless sensor networks, frequent data access will waste a lot of energy if the distances between the storage nodes and source nodes, or the data consumer nodes are too far from the storage nodes. This dissertation analyzes the strategy of local-based storage, points out the LS strategy has the advantage that its communication spending is relatively small in the storage procedure. On this basis, a data storage scheme called distribute index data storage scheme, which integrate both the advantages of local storage and index structures is proposed. By constructing multi-hop connected dominating set, this algorithm makes data to be stored on storage nodes, which have fixed number of hops away from the source node, the distances between the index nodes and the storage nodes are within a fixed number of hops accordingly. This scheme reduces the energy consumption in the period of data transmission. Moreover, the scheme makes the storage and index nodes to be evenly distributed within the networks, and provide timely response to queries. At the same time, the algorithm considers the requirements for load balancing and system robustness.We transforms the research of data acquisition and storage in wireless sensor networks into specific application. This dissertation design the bridge structural health monitoring (SHM) system based on wireless sensor networks, the overall design of the bridge structural health monitoring systems and the detailed design of subsystem are given in this dissertation. With regard to the weak signal acquisitions of sensors In wireless sensor networks, we study the signal acquisition and signal processing about vibration signals and strain signals. Concerning the characteristics of the bridge structural monitoring, this dissertation presents a data storage algorithm based on the cluster hierarchy. Finally, we achieve the bridge structural health monitoring system, deploy and test it on an actual bridge. The test has proved that it can achieve good results.