Research On Controller Design And Network Optimization Method In Cloud Control System

The cloud control system is the product of the combination of network control systems,Internet of Things and cloud computing technologies.The powerful computing power provided by the cloud computing platform can meet the high quality and real-time control requirements of the massive data control system.At the same time,the cloud control system puts forward new requirements for the controller,and the system inevitably has problems such as delay and unstable performance.Therefore,this paper studies these issues in cloud controller design and virtual network optimization methods.This paper first designs a cloud controller based on the structure of the cloud control system.The controller includes a variety of control algorithms for the user to select online,and can accept multiple client connections at the same time,and the client can set the parameters of the controller online.At the same time,the website is deployed on the cloud server,the users can access the current control status anytime.The cloud controller is portable,common,network access and easy to operate.Secondly,because the virtual network provides the resource platform for the cloud platform,the reliable network performance can reduce the delay caused by the cloud controller to obtain the cloud platform service and computing.In order to efficiently map the virtual network to the shared underlying network and reduce the mapping time consumption,this paper establishes a virtual network mapping problem model,and proposes a virtual network mapping algorithm based on multi-objective particle swarm optimization to solve the model.The cross-operator idea of genetic algorithm is integrated into the algorithm,and the non-dominated sorting and crowding distance sorting methods are effectively combined to accelerate the optimization speed of multi-target population.Simulation experiments show that the algorithm improves the optimization speed and accuracy,and achieves a better balance of the underlying network resources.Finally,in order to improve the flexibility of network resource allocation,this paper studies the elastic optical network as the underlying network architecture of the cloud platform.The integer linear programming model of RSA problem in optical network is established,and the traditional heuristic RSA algorithm is analyzed and improved.On the basis of ensuring network performance,time consumption is reduced.At the same time,the traditional RSA algorithm has a huge time consumption problem for large-scale network transmission.This paper firstly proposes a spectrum allocation algorithm based on particle swarm optimization for large-scale real-time network transmission,and solves the spectrum allocation of optical network with multi-peak characteristics.The problem is improved accordingly,and the effectiveness of the algorithm is proved by simulation.