Design And Optimization Of A Linux-based Real-time Decoding System Software

Towards building a safe and reliable information system industry chain in the rapid informatization development period,our country is paying more and more attention to the independent and controllable information system,especially in the field of satellite remote sensing applications.Limited by the transmission bandwidth,the massive monitoring data captured by the satellite cameras needs to be compressed before being transmitted down to the ground system.This compressed data should be decoded and processed by the ground system in real-time then.The transport rate of the image compression data would be always changed due to the different imaging modes,image sizes,and compression ratios.Thus,it would be difficult for the ground decoding system to decode this data in real-time.The existing ground decoding system software in use is developed based on the Windows system,which cannot satisfy independent and controllable requirements.This thesis studies the software development under the homebred Linux-based operating system,designs and constructs a fully independent and controllable real-time satellite image decoding system software.Through the efficient control of the image compression data and the hardware decoding card,the problem of large rate variation of compressed data is solved,and the realtime and stable decoding of the compression data is implemented.Firstly,the requirements of the decoding system,composition structure and main difficulties of the decoding system software are analyzed in detail.Secondly,the design scheme is developed according to the requirements of the decoding system software under Linux,and four processing modules with high cohesion and low coupling are designed to implement the real-time decoding system software under Linux.After that,in order to further improve the real-time performance of the system,the source of processing delay of the decoding system is analyzed,and the transmission time model of each buffer of the decoding system software is established.Based on this,the optimal delay optimization scheme is designed to make the decoding system maintain a very low delay（150ms for processing scan data and 100 ms for processing gaze data）while the data throughput rate can reach 1.5Gbps.Finally,to improve the security and stability of the system,the threadsafety of the decoding system software is analyzed and demonstrated.The coupling degree of the system is reduced and the reliability of the system is improved effectively since the threads with the high coupling degree in the software are optimized.In this thesis,the design and optimization of the decoding system software under Linux are studied to implement a fully independent and controllable real-time decoding system software,which satisfies the practical application requirements.Results proposed in this thesis improve the domestication degree of the decoding system software in the field of satellite remote sensing,and would effectively promote the development of domestication software ecology.