The Design And Implementation Of FPGA Bigdata Visualization Platform Optimized By Ant Colony Algorithm

Big data has now become a hot topic.Not only are various e-commerce platforms able to accurately push target commodities through big data behavior analysis of users,but also various big data platforms report the current epidemic situation through the analysis of the COVID-19 situation.The application of big data in these scenarios makes more and more industries hope to make more contributions to the industry by using big data.In the context of "new engineering" in college education,the reform of experimental teaching is a top priority.Facing the problem that the evaluation system of students' experimental quality is not comprehensive,big data can play its role to help the teachers evaluate the experimental quality of a student in a comprehensive and multi-dimensional manner.This paper designs a big data collection and analysis platform for FPGA.The platform collects logs of students on FPGA,describes students' use of "portrait" by big data analysis and calculation,and "portrait" reflects various situations of students' operation of FPGA experiment.Teachers can understand each student's learning of FPGA experiment course by using the platform.The design of the platform mainly includes three parts: the first part is how to realize the FPGA data collection in each student's hands and send the collected data to the platform;the second part is how to efficiently analyze and calculate these data;the third part is how to show the analysis results clearly and effectively to the user.In order to complete these contents,this paper has Done the following work:(1)A data acquisition board is designed for FPGA development board,which can collect some key data of FPGA,and realize data transmission between acquisition board and platform through network communication technology.(2)This paper investigates the current big data framework,formulates the technical selection of Hadoop and Spark,and optimizes some shortcomings of Hadoop application in this project.Based on the high efficiency of ant colony algorithm in solving combinatorial optimization problems and the characteristics of Hadoop scheduling,this paper puts forward the optimization scheme of improving Hadoop task scheduling by ant colony algorithm.(3)Build FPGA big data visualization platform,through the project requirements,functional design,system module design and other aspects of the detailed description of the platform.(4)Design the system test,through the test results to verify the ant colony scheduling algorithm on the design platform performance optimization effect and the completion of all the target functions.Platform has achieved the expected design requirements of the project.