Research On Data Collection Protocol For Wireless Sensor Networks Based On Data Persistence In Disaster Scenarios

Wireless sensor networks(WSN)have the inherent characteristics of selforganization,without human duty and rapid deployment,and play an increasingly important role in disaster as fires,landslides and floods.In such a disaster environment,sensor nodes is facing random destruction regional,which greatly increases the risk of data loss.Data persistence is an important indicator of test data availability.In the disaster environment,the destruction of the network node directly leads to the loss of the data.At the same time,the transmission delay of the wireless sensor network has greatly increased the risk of data loss.Therefore,how to improve the data persistence in disaster scenario is a very meaningful and challenging research.A number of domestic and foreign scholars focus on improving the persistence of data in disaster scenarios.Among these techniques,GC(Growth Codes)encoding technology and related follow-up study are the most representative.However,the sink frequently collects redundant data and seriously affects the performance of the data collection due to the use of a completely random data transmission way in GCP,which make data collection is not efficiency especially in the latter part of the data collection and sparse networks.Based on the research of GC data collection protocol,this paper focused on the problems described above made following studies:First of all,to solve the poor data collecting efficiency of GC,we propose a dynamic selection mechanism based communication node for data collection protocol(NGCP).NGCP control node communication show the following trends,early approximate random selection of communication to ensure the robustness of the network,then the communication probability dynamically add different stages of nodes to improve data collection efficiency edge.In order to optimize the realization of the above-mentioned trend,NGCP designed perception series table,new communication node probability function and genetic algorithm.Experiment results show that NGCP effectively imporves the collecting efficiency on the base of data persistence reliability.Secondly,in order to solve the problem of low performance of GCP caused by the random flow of data packets,we propose a data collection protocol based on hierarchical packet fast fusion strategy(DGCP).DGCP assigns to all network packets lightweight design data packet routing information,and controlled the dynamic probability function to improve data collection efficiency.Experiment results show that NGCP effectively imporves data persistence reliability.Thirdly,Combined with the advantages of NGCP and DGCP,the algorithm is improved,and we propose a data collection protocol based on the dynamic selection mechanism of communication nodes and the fast fusion strategy of hierarchical data packets(MGCP).Through performance evaluation,MGCP has better performance than NGCP and DGCP in different types of disaster environment,and has better adaptability to different disaster environment.Research on wireless sensor network transmission characteristics of their networks in disaster scenarios to improve data collection,which promotes the development of related technologies of wireless sensor has a certain role.