Research And Design Of Using Deep Reinforcement Learning To Improve Internet Of Things Network Flux

In recent years,the Internet of things business such as intelligent transportation,smart home,environmental monitoring,public safety and industrial monitoring has developed rapidly.The terminal equipment in the Internet of things has increased explosively,and a large number of Internet of things devices have brought massive data transmission traffic.Traditional traffic scheduling strategies and routing schemes,such as open shortest path,can not route the data traffic according to the real-time status information of the network,and it is difficult to ensure the service quality of Internet of things business.Therefore,it is necessary to propose an intelligent Internet of things traffic control scheme to realize the load balance of Internet of things and improve the network throughput.Deep reinforcement learning has great potential in realizing intelligent control.However,the distributed nature of existing network architecture is not conducive to the application of deep reinforcement learning technology in intelligent control of network.The emergence of software-defined network technology solves this problem,and its advantages of separation of control and forwarding and centralized control are conducive to applying deep reinforcement learning to the Internet of Things to realize real-time intelligent routing scheduling of data traffic in the Internet of Things.In this paper,the problem of load balancing of Internet of Things network traffic is studied.Combined with deep reinforcement learning and software-defined network technology,a real-time intelligent routing scheme of Internet of Things data traffic is proposed.The specific research work is as follows:(1)This paper analyzes the existing network routing schemes and their existing problems.The load balancing problem of Internet of things is systematically analyzed and modeled.Combined with deep reinforcement learning and software defined network technology,SDN intelligent routing algorithm is proposed.The software defines the network technology to obtain the network status information in real time,and the SDN intelligent routing algorithm calculates the real-time intelligent optimal routing decision based on this network information,so as to realize the load balance of the Internet of things and improve the network flux.(2)This paper studies the data flow fluctuation caused by terminal equipment accessing the network in the Internet of things,analyzes and models the network data flow fluctuation,and puts forward the threshold adjustment mechanism of link bandwidth utilization.Based on this mechanism,the terminal equipment accessing the network routing algorithm is designed.When the terminal equipment accesses the network,the mechanism adjusts the bandwidth utilization threshold of the network area link accessed by the terminal.The terminal equipment access network routing algorithm formulates a reasonable data traffic routing scheme according to the changed threshold in the network and the network state information to solve the problem of data traffic fluctuation in the network.