| With the update and development of GPS in America,modernization of GLONASS in Rusia,operation of the Galileo in Europe and improvement of BDS in China,the four global navigation satellite systems has developed tremendously in the space segment.As the same time,the scale of GNSS network is bigger and bigger in the ground segment,and thus the GNSS data quantity increases exponentially.Both of these indicate that GNSS data has entered the era of big data.However,massive GNSS data not only brings opportunities,but also provides rich computing resources,which means great challenges in computing and management.Therefore,how to efficiently store massive GNSS data in real time,improve the efficiency of reading and downloading,and alleviate the contradiction between massive data and limited computing,storage capabilities has become an urgent problem to be solved nowadays.In view of this,this paper has carried out research on the above issues,the major contents and achievements are as follows:1.The current development status of massive GNSS data is analyzed from the aspect of space segment,ground segment and data quantity.Available satellites and frequencies increase significantly in the space segment,while the construction and upgrade of the ground station base station network is more conscious in the ground segment,and thus the GNSS data quantity increases exponentially.Then,the storage and management method of massive GNSS data are introduced in detail,which means centralized storage method cannot meet the needs of massive GNSS data.Therefore,a technical route for GNSS data storage and management based on distributed file system in Hadoop architecture is proposed.2.Build Hadoop server cluster.Through the study of cloud computing and large data,the opening source cloud computing and storgage framework Hadoop is expounded,especially the most important components: HDFS and MapReduce.Then the paper has tested the reliability,fault tolerance and expansion performance of the cluster,while set up a Hadoop small server cluster successfully.And the test results show that the Hadoop cluster has good security,reliability,disaster tolerance and scalability.3.In the research of GNSS small file upload storage strategy,according to the matter of the low processing efficiency of massive GNSS small files faced by HDFS,a new storage scheme is proposed.GNSS Small file merge based on GNSS file type,time and location information,and then enhanced index storage scheme by building a small file index table,that is,a categorized compressed index storage method.Through experiments,this scheme is compared with the original HDFS storage,HAR file scheme and SequenceFile serialization scheme.The experiment and analysis of the four schemes were carried out from three aspects: file uploading experiment,file reading and main node memory loss.The results show that this scheme is more effective in solving GNSS small file storage and management.4.The software of GNSS cloud disk system is developed and tested.Making use of theoretical and algorithmic results mentioned above and based on the analysis of system requirements,the GNSS data cloud disk system was developed ultimately,which realizes the cloud storage of GNSS files.The main functions of the system include: function main menu,login function implementation,file interoperability implementation,file lab cloud sharing implementation,GNSS data storage implementation,and GNSS positioning implementation.And part of the GNSS cloud functionality is implemented on the web side of the GNSS cloud disk. |
PDF Full Text Request