Design And Implementation Of Efficient Transmission System Based On ZYNQ Platform

With the vigorous development of emerging technologies such as big data and cloud computing,various application tasks emerge in endlessly.In order to meet the needs of diversified processing of different application tasks,heterogeneous platforms have become a research hotspot.ZYNQ heterogeneous platform integrates ARM and FPGA on one chip.It not only retains the software programmability of CPU,but also expands the hardware programmability of FPGA,so it is widely used.However,for heterogeneous platforms composed of different processors,if the scheduling management and data processing of different application tasks are realized by CPU,it will have a certain impact on the efficient transmission of heterogeneous platform systems.Therefore,this paper studies the efficient transmission of heterogeneous platforms.Based on the complexity and differential processing requirements of multi application tasks in ZYNQ platform,this paper puts forward different solutions,and uses the advantages of different processors to achieve the purpose of efficient transmission of heterogeneous platforms and improve the overall performance of the system.This paper studies from the following aspects:(1)TOE design based on HLSAiming at the problems of large CPU processing burden,low occupancy and poor flexibility and efficiency of traditional register transfer level RTL design in communication protocol processing,HLS high level synthesis technology is adopted to realize TOE design,improve the flexibility and efficiency of algorithm design and reduce the CPU processing burden.The experimental results show that compared with the CPU execution efficiency,the TOE system realized by HLS technology obtains higher throughput,improves the performance by 8.9%,and optimizes the average area of resource occupation by 10.3%.(2)Multi application task traffic management algorithmAiming at the problem of packet loss caused by multi application task data burst transmission in FPGA on heterogeneous platform,based on the analysis of several classical traffic management algorithms,a traffic management RFCF algorithm suitable for FPGA is proposed and implemented.The remaining resources are fed back to the sender by using control characters,and the sender adopts different data priority transmission strategies according to the resource margin to ensure that the burst data under multi application tasks can be processed in time.The experimental results show that the packet loss rate of RFCF algorithm proposed in this paper is 14.9% lower than that of rate based traffic management algorithm,and the resource occupation is optimized by 14%.(3)Design and implementation of video target detection system based on ZYNQConcretize the typical application tasks,build a video target detection system based on ZYNQ,apply TOE design and RFCF algorithm to the target detection system,and then debug and analyze the system.Through the test and analysis of hardware environment and software environment,the results show that TOE and RFCF algorithm module can effectively improve the system transmission performance and meet the functional requirements of the system design.