Design And Implementation Of Network Data Transmission And Fast Processing Architecture Of OpenSAR

SAR image dataset is the basis of SAR image interpretation,but the founding of SAR image dataset of large quantities is a challenging work,which is an urgent problem for us to tackle.In order to promote research in the field of SAR image interpretation,this paper introduced to build a platform,OpenSAR,which is a collection for a great quantity of SAR data as well as a testing platform for SAR image interpretation algorithms.OpenSAR aims two,the one is to construct large quantities of SAR image dataset,and the other is to implement an algorithm library for SAR image algorithms.With the guidance of these main aims,this paper described the problems we may face and analyzed the technical difficulties,and then points out the importance and necessity of research on large SAR image network transmission and the faster calculation architecture of SAR image.The foundation of OpenSAR will be expected to promote China's technological advances in the field of SAR image interpretation.Below are the main innovations,as well as key contributions: ·Detailed design of Architecture for OpenSAR platformThe platform can be divided into presentation layer,functional layer and data service layer.This paper described the structure of OpenSAR platform,both in software and hardware,as well as components which construct OpenSAR,and give detailed design of each layer.This paper focused on some key point technologies such as network data send/recv,concurrency in the service provider,data service assurance.·Detailed research and design of SAR image network transmission module SNTMThis paper gived detailed discussion of SAR image network transmission model based on data fragmentation,and described the technique such as data fragmentation,breakpoint continuingly uploading and fragment reassemble,etc.Determined the two key parameters via optimizing experiment.On the basis of theoretical research finally designed SNTM.·Detailed research and design of SAR image distributed computing framework RTPThis paper gives detailed discussion of SAR image distributed computing framework based on Hadoop,and described the operating mechanism of Map/Reduce Module.Straight Hadoop calculations may lead to the loss of key SAR image features.To solve this problem this paper introduced AIT module and determined the key parameter via optimizing experiment.With all the above finally designed RTP.