面向无人机网络的基于随机森林的协作邻居发现协议研究

Research on Cooperative Neighbor Discovery Protocol Based on Random Forests for UAV Networks

随着无人机的迅速发展,对无人机组网的研究也成为了热点。邻居发现作为组网完成的第一步,发挥着十分重要的作用。首先介绍了无人机网络的场景和无人机的结构,然后提出了基于随机森林的协作邻居发现协议,将三次握手机制和随机森林分类器应用到邻居发现协议中,通过将自身邻居无人机节点推荐给目标无人机,提高了邻居发现的速度。通过仿真,从四个不同方面比较分析了不同协议之间的性能差异,结果表明,RFC-CND协议具有更快的发现速度和更高的邻居发现率。

Neighbor discovery is the first step of networking and plays an important role.This paper first introduces the UAV network scene and UAV structure.Then this paper proposes a cooperative neighbor discovery protocol based on random forest,and applies the three-way handshake mechanism and random forest classifier to the neighbor discovery protocol,and improves the neighbor discovery speed by recommending the neighbor UAV node to the target UAV.This paper compares and analyzes the performance differences among different protocols from four different aspects through simulation experiments.The results show that RFC-CND protocol has faster discovery speed and higher neighbor discovery rate.