车联网中基于K均值聚类的拥塞控制方法

针对十字路口下车辆密度过大时造成的车联网拥塞问题,提出基于K-means聚类的车联网拥塞控制方法。在IEEE802.11P协议模型的基础上,引入处理高并发数据的拥塞控制模块,利用网络层分簇时V2V相对距离值作为K-means聚类算法帧分类的相似度区分值,与VANET的网络标准参数建立拥塞控制方法(KCC),避免同一类车辆节点交互不同步和高密度节点情况下网络拥塞概率过高等问题。仿真结果表明,KCC与CSMA/CA相比有较好的平均时延、平均吞吐量、丢包率、冲突概率等性能。

Congestion control method based on K-means clustering in VANET

Aiming at the congestion problem of vehicles caused by excessive vehicle density at intersections,a vehicle-based congestion control method based on K-means clustering was proposed.Based on the IEEE802.11P protocol model,a congestion control module for processing high concurrent data was introduced.The V2V relative distance value was used as the similarity value of the K-means algorithm frame classification when the network layer was clustered,and congestion was established with the VANET network standard parameters.The control method(KCC)avoided problems such as the unsynchronized interaction of the same type of vehicle nodes and the high probability of network congestion in the case of high-density nodes.The simulation results show that KCC has better average delay,average throughput,packet loss rate and collision probability than CSMA/CA.