Software defined networking (SDN) architecture is easily manageable, flexible, and dynamic where controller selection plays a major role. In SDN, the controlling functions are handled by control plane and forwarding functions are handled by data plane. In general, the centralized controller supervises the control plane, and it communicates with the data plane using “Open Flow” (south bound interface). Moreover, OpenFlow based controllers such as POX, Floodlight, Beacon, and RYU have different features which suit different applications in real time. In this proposed work, a multi-criteria decision making method named entropy based TOPSIS method is used to select the controller for load balancing in control plane. Initially, four controllers are selected and evaluated over tree topology with different parameters such as throughput, delay, response time, message cost using the network emulator called Mininet. Then, entropy based TOPSIS method is used to rank the controllers where Floodlight is chosen as the most excellent one based on our criteria.

中文翻译：

---

基于熵的 TOPSIS 方法用于软件定义网络中的控制器选择

软件定义网络 (SDN) 架构易于管理、灵活且动态，其中控制器选择起着重要作用。在SDN中，控制功能由控制平面处理，转发功能由数据平面处理。一般情况下，集中控制器监督控制平面，它使用“开放流”（南向接口）与数据平面通信。此外，基于 OpenFlow 的控制器，如 POX、Floodlight、Beacon 和 RYU，具有不同的特性，可以实时适应不同的应用。在这项建议的工作中，一种名为基于熵的 TOPSIS 方法的多准则决策方法用于选择控制器以在控制平面进行负载平衡。最初，选择四个控制器并通过树形拓扑评估不同的参数，例如吞吐量、延迟、响应时间，使用名为 Mininet 的网络模拟器的消息成本。然后，使用基于熵的 TOPSIS 方法对根据我们的标准将 Floodlight 选为最优秀控制器的控制器进行排名。