Wireless Sensor Network Data Storage And Query Optimization Strategies Based On RCFile

With the development of wireless sensor network technology, it is widely in many fields,like the mili-tary and national defense, industry and agriculture, urban management, biomedical, environmental moni-toring, disaster relief, counter-terrorism and anti-terrorism. Data-centric is the most important feature ofwireless sensor network, so the key point is how to store and transmit the data effective perceived by nodes,which directly affects the reliability and efficiency of the data access. Since the sensor network covering awide area, has a large number of nodes, and monitoring data every minute, so the nodes need to receive andprocess large amounts of data. Because of the energy and the storage capacity of each node is limited, manynodes are died and the data lost caused by the accumulation and redundancy of large amounts of da-ta.Storing and transmitting data resonable and effective, which could reducing energy consumption duringdata transmission, improving the efficiency of the whole network and prolonging the life of the sensor net-work.To solve this problem, this paper first gives a detailed discription about data storage and research ofwireless sensor networks at home and abroad. Firstly, the domestic wireless sensor network data storageand query algorithms line classification elaborate, and the advantages and disadvantages of them gives acorrelation analysis. Then proposed an efficient large data storage structure-RCFile, based on RCFile pro-pose a suitable sensor network data storage algorithm. The following four areas are the main contents ofthis paper:1. The article described the problem of the data storage problems exist in wireless sensor net-works.There are there solutions strategies:(1) external storage;(2) local storage;(3) data-centric storage.In this paper, there will be a disscution about the three data storage strategies and algorithms in detail, thengiven the advantages and disadvantages of these three storage strategies.2. There is a description about characteristics and trends of big data in this section. In the storage ofbig data currently, they are row-store, column store and PAX store. When dealing with huge amounts ofdata, there are a lot of contention in academia about which one could deal with data more integrity, securityand reliability. This paper discusses the storage format and structure of the model about the three storagestyles. 3. This section first studied the characteristics of wireless sensor network, then discussing the similari-ties and differences between big data and data in WSN. Based on RCFile, a data storage algorithm is pro-posed. RCFile is an efficient data storage structure in big data. While after changing the structure of WSN,the structure of RCFile will applied into WSN. This will improve the utilization of space, reducing theenrrgy consumption, finally extending the life of nodes and increasing the overall network performance.4. Based on RCFile, a data query algorithm is proposed. Combineding with the Distributed processingand parallel computing, using the the advantages of RCFile storage, this could reducing the redundant dataquery process.During the querying, lazy decompression will be used. This could skipp unnecessary col-umns inquiries, to speed up the query response time. In the data gathering and loading process, combinesthe advantages of the row-store, reducing network overhead. Compared with other row-store queryprocessing algorithms, the core of the algorithm is to use the characteristics of the row storage node selec-tion queries.This not only reduce energy consumption of query flooding across the network, also decreasethe query response time.