SDN下分布式负载均衡博弈算法研究

Research on Distributed Load Balancing Game Algorithm under SDN

随着互联网产业中数据流量的不断增长,数据中心服务器的处理能力和内存访问速度也不断提高。为解决升级硬件层面成本高、可扩展性差的问题,本文提出了一个基于分布式马尔可夫博弈的自适应负载均衡系统。系统利用SDN的全局视图与动态控制能力,结合分布式马尔可夫博弈模型,实现多VNF实例的智能调度与资源优化。仿真结果表明,相比于传统的集中式负载均衡算法,本文提出的算法在提高网络可扩展性及降低网络能耗方面具有明显优势。

With the continuous growth of data traffic in the Internet industry,the processing capacity and memory access speed of the data center server are also increasing.To address the issues of high cost and poor scalability in upgrading hardware,this paper proposes an adaptive load balancing system based on distributed Markov game theory.By utilizing the global view and dynamic control capabilities of SDN,combined with a distributed Markov game model,intelligent scheduling and resource optimization of multiple VNF instances can be achieved.The simulation results show that compared to traditional centralized load balancing algorithms,the algorithm proposed in this paper has significant advantages in improving network scalability and reducing network energy consumption.