A Study Of Distributed Storage Strategies Based On Fountain Codes In Wireless Sensor Networks

Wireless network is a typical distributed storage system. It is data-centric and self-organized, so data storage and data collection is very important. The sensor nodes are often deployed in relatively hash or restricted environment, they face unpredictable fault or die which seriously impact the network data integrity and availability. In wireless sensor network, the methods of data storage directly impact the efficiency of data collection. So how to ensure the reliable storage and effective collection becomes a worth research.Recently, the applications of network coding technology in WSN is more and more extensive, it breaks the network node the traditional processing method of simply copy and forward. In WSN, the introduction of the fountain code technology can enhance the data persistence. But the traditional distributed coding strategies may cause serious "cliff effect" in decoding process. That is to say, only few source data can be recovered before sufficient encoded packets received. Moreover, nodes may fail due to the lack of energy or the influence of the switching of external environment. Such failures may concentrate in a region which can severely reduce the decoding efficiency of the persistent data.In this paper, we focus the problems of network data reliable storage and low data collecting efficiency, and made the following research work:(1) This paper broke through the assimilation characteristics of behavior and data distribution in homogeneous network, designed a class of prioritized coding degree distribution strategy and proposed a distributed storage algorithm with high efficient decoding PLTCDS. This strategy makes the encoded data present a trend that their degree distribution increases by simply the predefined sink node broadcasts a signal, and the persistent data packets can be submitted to the sink node in order to meet demand decoding. Experimental results show that PLTCDS can enhance the data collection and decoding performance while data persistence is assured.(2) In this paper, we consider the empty storage problem, and designed a simple storage counter mechanism to ensure nodes effective storage, to improve the efficiency of data collecting. Compared to the original strategy, this mechanism can make full use of survived nodes in network.(3) In view of the actual network scene, this paper extends PLTCDS algorithm to the edge collection scene, design another kind of prioritized LT codes based distribution storage algorithm PLTCDS type Ⅱ. The central node broadcasts a beacon to stimulate the nodes to form the network with degree distribution priority, constructs network with degree level, designs a code distribution, and optimized the encoding data allocation policy. Experimental results show that PLTCDS TPYE Ⅱ can enhance the data collection and decoding performance while data persistence is assured.In wireless sensor network, especially in the scene of disaster and nodes fragile, the research of how to improve the storage reliability and data collection is very important. And it can make a significance to promote the development of wireless sensor network technology.