活跃度感知的社交车辆分簇算法

Activity-aware social vehicle clustering algorithm

为了解决车联网(IoV)中因车辆高速移动和拓扑结构多变导致的车辆间数据传输链路不稳定甚至中断的问题,提出一种活跃度感知的社交车辆分簇算法.在簇头(CH)筛选过程中,考虑由相对加速度、速度和相对距离构成的移动相似性分值以及由兴趣相似度定义的社交相似性分值,加权求和得到车辆相似性分值.利用基数排序算法排序并筛选出分值最高者作为簇头候选者(CHc),保证集群的稳定性.引入由车辆历史数据处理量和车辆请求资源次数构成的活跃度的概念,通过对其进行判断,从簇头候选者中筛选出真正有社交意愿和能力的簇头,提升簇内亲密度.使用OMNet++平台进行仿真,结果表明,与传统算法相比,采用所提算法,能使得集群在保持稳定性的同时,亲密度有所提升.

An activity-aware social vehicle clustering algorithm was proposed, in order to solve the problem of instability or interruption of data transmission link between vehicles due to the high mobility of vehicles and the changeable topology in the Internet of Vehicles(IoV). In the cluster head(CH) selection process, the mobility similarity score composed of relative acceleration, speed and distance and social similarity score defined by interest similarity were considered, then were weighted and summed to obtain similarity score. The radix sorting algorithm was used to sort and select cluster head candidates(CHc) with highest scores, which ensured the stability. Activity degree consisting of the amount of historical processed data and the number of requests for resources was introduced.By measuring it, the CH with real social willingness and ability was selected from CHcs, and this increases the intimacy within clusters. Simulation results on the OMNet++ platform show that compared with traditional algorithms, the proposed algorithm maintains the stability of the cluster while increasing the intimacy.