融合社会关系的机会网络有效数据转发策略

Effective Data Forwarding Strategy Integrating Social Relationships in Opportunistic Networks

机会网络是一种具有延迟容忍网络特征的移动自组织网络。随着各种具有短距离通信功能的便携式移动设备的迅速普及,具有社会性的机会网络应用场景增多。机会网络中具有类似于节点聚集的现象,节点表现出的某些特征呈现出了社区结构的特性。然而,现有的路由算法没有考虑到节点社会性对网络中消息传递的影响,其传输成功率较低且造成大量的时延和网络开销。针对该问题提出了一种基于社区和社会性的数据转发机制,根据节点间的社会关系将网络划分成若干个社团结构,通过分析社区内节点的重要程度删除一些低效节点,并对这些社区结构进行再收缩,使社区结构紧密,提高传输效率。仿真结果表明,该算法相对Spray and Wait算法、PRo PHET算法和SCR算法(effective social relationship measurement and cluster based routing in mobile opportunistic networks),具有较高的传输成功率和较低的传输延迟。

Opportunistic network is a kind of Ad hoc network and it also has characteristics of delay tolerant networks. With the rapid popularization of portable mobile devices which have short distance communication functions, there are more application scenarios for social-based opportunistic network. Thus opportunistic network also has a phenomenon similar to node aggregation. Some features displayed by nodes in the network exhibit the characteristics of the community structure. However, existing routing algorithm doesn..t consider the social characteristics of nodes, which causes a low packet delivery ratio and high ratio of transmission delay and routing overhead. To solve these social-based routing problems, this paper proposes a data forwarding method under fully consideration of community and social features. According to the social relationships among nodes, the network can be divided into several community structures. This paper addresses a method to reduce the community size, using this algorithm to delete nodes which are inefficient, and makes the community structure centralized to reduce the energy consumption of useless nodes. Through a series of actions, this paper can improve the transmission efficiency. Simulation result shows that this algorithm has higher delivery ratio and lower transmission delay compared with Spray and Wait algorithm, PRoPHET algorithm and SCR(effective social relationship measurement and cluster based routing in mobile opportunistic networks) algorithm.