人工蜂与K-means混合算法在VANETs的应用

针对车载自组织网（Vehicular Ad Hoc Networks，VANETs）拓扑结构经常变化导致通信链路容易断裂而通信质量不可靠的问题，将人工蜂与K-means混合算法应用在VANETs中。在成簇阶段，该混合算法利用人工蜂算法较强的全局搜索能力确定初始聚类中心，代替传统的K-means对初始聚类中心的选择，这样就消除了K-means对随机初始聚类中心的依赖。在簇头选取阶段，类内具有最小的速度方差以及到其他节点最小平均距离的车辆节点被选择为簇头。在簇的维护阶段，当最优节点即簇头有变化时，次优节点被选为临时簇头，直至更新为最优节点的簇头信息。为测试该混合算法的性能，将其和PSO与K-means混合算法、经典K-means算法进行实验对比，结果表明，该混合算法能够更加稳定VANETs通信链路，具有更高成簇质量和更高通信质量。

Application of the Hybrid Algorithm of Artificial Bee Algorithm and K-means in VANETs

The VANETs topology often changes and the communication link is easy to break and the communication quality is unreliable.To solve this problem,a hybrid algorithm of artificial bee and K-means is applied to VANETs.In the clustering stage,the hybrid algorithm uses the strong global search ability of the artificial bee algorithm to determine the initial clustering center,instead of the traditional K-means selection of the initial clustering center,which eliminates K-means'' reliance on random initial cluster centers.In the cluster head selection stage,the vehicle nodes with the smallest speed variance and the smallest average distance to other nodes are selected as cluster heads.In the cluster maintenance stage,when the optimal nodes,that is,cluster heads,changes,the sub-optimal nodes are selected as temporary cluster heads until the cluster head information of the optimal node is updated.In order to test the performance of the hybrid algorithm,experiments were carried out to compare PSO with K-means hybrid algorithm and classical K-means algorithm.The results show that the hybrid algorithm can stabilize the VANETs communication link more stably,have higher clustering quality and can improve communication quality.