Design And Implementation Of High-Performance And Highly-efficient Location Information Management Software In Internet Of Vehicles

In recent years, researches of Internet of Vehicles become more and more perfect, the large-scale application of Internet of Vehicles is yet to come. This article designed a Web service software based on JavaEE that can be applied to the Internet of Vehicles. This system provided different characteristic services based on location information according to different types of users. The location information is collected from the vehicles by the cellular network. The system has three user types:administrators users, roadside businesses users and the vehicle users. The administrators can obtain the location information which vehicle upload through the cellular network and manage vehicle location information.Roadside businesses users can send their own promotional advertising information and so on to the users of passing cars and chat with vehicle users through the cellular network. The vehicle users upload their own position information through a cellular network. The upload information includes position, velocity, frequency of beacon in DSRC(Dedicated Short Range Communications), the packet loss rate.Then the background server gathers the uploaded location information and predicts the quality of wireless network by DSRC, using a machine learning algorithm.Then the server returns a suggest frequency of broadcast to the vehicles.Finally the vehicles brocast the beacon by the frequency given.This algorithm can reduce the packet loss rate and end-to-end delay in V2V communication by DSRC and then improves the performance of location updating. This paper also researches how to use Redis and Epoll to optimize the web application in the Internet of Vehicles, shorten the response time of the system, improve the server's concurrent processing ability, and provide reliable services.In this paper, the main completed works are as follows:(1) This paper introduces the open source frameworks used in software, analyzes the requirements of vehicle location information management system, and designes different location based services for different user types.(2) The vehicle location information management software is written based on Spring, SpringMVC and Mybatis development frameworks, and the logic flow of the software service is optimized. This paper enhances software's response speed in the Internet of Vehicles, and the concurrent service capability of the vehicle position information management system is improved by Redis and Epoll.(3) In this paper, NS-3 simulation results are facilitated as training set. The LR, Naive Bayes, SVM and C4.5 decision tree are used to judge whether the vehicle's DSRC V2V wireless communication environment has the risk of network congestion. The best training model is selected and applied to the application service of the system, and the quality of vehicle self-organizing network is improved by returning the recommended frequency of beacon and the performance of location updating between vehicles is improved. This scheme can be regarded as a cross-laye scheme for performance optimization of bottom location update by machine learning in application layer.(4) The function test and stress test are carried out on the login module, location management module, message management module and network quality module. The test results show that the system can manage the location information of vehicles and provide related services by using location information. The service time can meet the requirement of the Internet of Vehicles, and can cope with the high concurrency scenarios.