Research And Implementation Of Rate-Constrained Service Planning And Scheduling

In order to meet the higher requirements of complex systems for data transmission bandwidth,real-time,and reliability,deterministic networks such as AFDX and TTE emerged as the times require,and have been widely used in aerospace electronic control systems and automotive networks.Rate-Constrained(RC)communication service has the advantages of flexibility and dynamics,and improves service transmission efficiency through virtual links and fixed bandwidth.It has become one of the mainstream communication services of deterministic Ethernet.Because of the event-triggered characteristics of RC services,collisions will inevitably occur.Although the traditional scheduling method of RC services can get rid of the limitation of collisions,the maximum transmission delay and jitter of the resulting services are unpredictable.Quality of service(Qo S)cannot be controlled according to specific application scenarios.Therefore,the application of traditional RC services to network scenarios with controllable delay and planned jitter requirements is limited.Combined with laboratory projects,this thesis focuses on the research and implementation of Rate-Constrained business planning and scheduling.This thesis firstly introduces the research background of Rate-Constrained services and scheduling methods,as well as the research status at home and abroad.Secondly,the basic concepts and Qo S indicators of RC services are summarized.Combined with the characteristics of RC services in airborne networks,a planning and scheduling strategy for RC services based on time slices is proposed.By planning the time slice of the RC service in advance,a scheduling table can be formed.The composition and working principle of the schedule table are described in detail.At the same time,by analyzing the performance of the scheduling strategy,it is shown that the strategy can effectively control the fluctuation range of service delay and jitter,and enhance the certainty of data transmission.Third,combining recursive and dynamic programming ideas,three RC service scheduling table generation algorithms are designed,which can plan the scheduling period and offset of each RC service according to the priority of service delay and jitter.Thereby generating a schedule table of RC service planning.Fourth,the three RC service scheduling generation algorithms are implemented in software using Python language.Combined with specific scientific research scenarios,the scheduling table generated by the software is applied to the RC service scheduling of the TTE end system.And the code design and implementation of each specific module in the scheduling unit are carried out using Verilog language.Finally,the simulation test environment and the board-level test environment are built respectively.Through simulation verification and FPGA board-level verification,it is proved that the scheduling strategy described in this thesis is reasonable and effective,and the functions of the designed scheduling unit is correct,stable and reliable.The expected design goals have been achieved.The innovation of this thesis is that,first,a planning and scheduling strategy for RC services based on time slices is proposed,which can effectively control the fluctuation range of delay and jitter,and enhance the certainty of data transmission.Second,three RC service scheduling table generation algorithms are designed,which can calculate the period and offset parameters on demand.It can well support the rate-limited business planning and scheduling strategy.Third,according to the three scheduling table generation algorithms,the required scheduling table is directly generated through the Python language,which makes the scheduling table planning faster and more convenient.Fourth,in the TTE end system,a planning and scheduling strategy for RC services based on time slices is introduced,and a scheduling unit that realizes the mixed transmission of multiple services is designed and implemented.The theory and engineering practice are combined to enhance the real-time performance and reliability of the TTE scheduling.