标签分割的软件定义飞行自组网控制器智能部署方法

针对当前软件定义飞行自组网中多控制器部署面临的负载不均衡和网络可靠性问题, 提出一种标签分割的控制器智能部署方法。该方法能够在不预设控制器数量的条件下, 根据控制器容量约束和网络结构, 输出最佳控制器数量及部署位置。基于节点自身特征和关联特征赋予节点标签, 根据标签完成控制域划分并通过布谷鸟搜索算法优化划分过程。在此基础上，考虑控制器平均时延、负载差异度和控制域时延波动的影响, 确定控制器部署位置。仿真结果表明，所提算法能在保证网络可靠性的同时有效减少控制器数量, 降低部署成本; 同时, 降低控制器平均时延和负载差异度, 保证各控制域间平均时延的相对平衡, 实现网络均衡。

Intelligent deployment method of software-defined flying ad-hoc network controller based on label segmentation

To solve the load imbalance and network reliability problems faced by multi-controller deployment in software-defined flying ad-hoc network, an intelligent deployment method based on label segmentation for controller is presented. This method can output the optimal number of controllers and the deployment location according to the capacity constraints and network structure of controllers without preset number of controllers. Based on the node's own characteristics and the related features, the node labels are given. The control domain is divided according to the labels, and the partitioning process is optimized through the cuckoo search algorithm. On this basis, the controller deployment location is determined considering the effects of controller average latency, load difference and control domain latency fluctuations. The simulation results show that the proposed algorithm can effectively reduce the number of controllers and deployment costs while guaranteeing network reliability. At the same time, it reduces the average controller latency and load difference, and ensures the relative balance of the average latency among the control domains, so as to achieve network balance.