Dynamic placement of multiple controllers based on SDN and allocation of computational resources based on heuristic ant colony algorithm

摘要

With the rapid development of the Internet and the explosive growth of network applications, traditional computer networks have ushered in unprecedented challenges and opportunities. To solve the communication delay between controllers and switches and the communication problem between controllers due to link failure in the network, this paper considers the delay between controllers, the delay problem between controllers and switches, and the reliability problem. It proposes a dynamic controller placement method based on delay and load optimization. A multi-objective optimization model based on link failure is constructed, and the multi-objective optimization problem with constraints is solved by improving the controller prevention algorithm with spectral clustering. Meanwhile, this paper proposes a resource allocation method based on task delay and reliability constraints to solve the problem of considerable task completion delay and wasted computational resources due to uneven resource allocation of edge servers. A model based on task latency and dynamic constraints is constructed, and a heuristic ant colony algorithm solves an adaptive allocation scheme for computing resources. The experimental results show that the proposed resource allocation algorithm weighs the delay between controllers, the communication delay between controllers and switches, and the reliability and gives a reasonable controller placement scheme and controller locations. The proposed resource allocation algorithm can optimize computing resources and reduce task completion delay.