Research On Data Storage Methods Based On Geographic Location In Wireless Sensor Network

With the continuous advancement of microelectronics technology, embeddedtechnology and wireless communications technology, people highly integratedinformation collection, data processing, wireless communications and other functionsin micro-sensor. Wireless Sensor Network, by the deployment of a large number oflow energy consumption, multifunctional sensor nodes in the monitoring area, formedby wireless means of communication, perceives, collects and processes objects in thecoverage area, and send the information processed to the needed users. This multi-hopnetwork system is mainly used in military, environmental science, health care, spaceexploration and other commercial areas.Wireless Sensor Network monitors real-world goals and produces large amountsof data after the micro-processor embedded within it has dealt, and passes to the enduser by wireless communication.WSN is data-centric network, data management anddata processing are the core technologies for it. In Wireless Sensor Network, the threemethods, external storage, local storage and data-centric storage, data-centric storagehas the best performance. Date-centric storage put the data in one or more sensornodes in the network according to event names generated by the node. All the sensornodes and sinks in WSN know the data center, data storage and query is not blind,which helps to reduce the energy consumption of storage and query. In muchdata-centric storage, the method based on geographic hash table is effective it isgenerally used based on a query-driven network. It effectively solves the packetflooding, and significantly reduces total energy consumption. However, in thedata-centric storage and GHT there exists load balancing and hot concentratedcommunication, handing node failure is also not perfect. This paper mainly studieshow to make better use of the limited energy when storage data, to maximize the network life cycle, and to enhance the robustness and survivability of wireless sensornetwork.On the basis of GHT, this paper presents an energy aware data storage methodbased on data gravitation. When nodes generate events, we get the data storagelocation L according to the predefined hash function, GPSR forwards to the nodelocating at L, that is, the home node. If there is no node at L, the node nearest to Lbecomes the home node. The nodes in the network record energy of their own. Whentheir energy reduces down to pre-set threshold, backup data to the replica node, thehome node turn to “sleep” state, that is, no longer store data. If nodes collect sametypes of data, and it needs to store data in home node which has slept, we store dadain the extended node according to calculating data gravitation. When there are queriesin the network, it calculates with the same hash function and sends to the home node.If there is replica node, then forwards the request to the replica node. We use thesimulation tool OMNeT++to test PRP algorithm in the GHT method and theproposed DGEA algorithm. The experimental results show that the proposed DGEAalgorithm can able to balance the energy load of nodes, effectively improve thenetwork life cycle, and has a certain degree of robustness and survivability.