BIRI:支持信息中心范型的BBO启发式MSN路由算法

BIRI: A BBO-Inspired MSN Routing Algorithm with Information-Centric Paradigm Support

智能移动终端的普及大大推动了移动社交网络(mobile social networks, MSNs)的发展.人类作为终端设备的载体具备频繁的移动性,导致网络拓扑的动态变化,并给MSN路由带来了时延长、投递率低、开销大等诸多难题.为提升路由效率,基于信息中心网络(information centric networking, ICN)以内容为中心的思想以及生物地理优化(biogeography-based optimization, BBO)算法,设计了一种高效的支持信息中心范型的BBO启发式MSN路由算法(BBO-inspired MSN routing algorithm with information-centric paradigm support, BIRI).首先,该机制基于重定义的社交度量——社会关系强度和中心度——使用BBO算法进行社区划分.其次,设计了内容聚集、数据缓存以及桥节点选取策略,支持高效的内容检索和访问.基于上述策略,提出了优化的社区间和社区内路由过程,缓解终端移动性对数据传输带来的影响.在机会网络环境(opportunistic network environment, ONE)中,仿真实现BIRI机制,并且与其他3种MSN路由机制从投递率、平均时延、网络开销比率3个指标进行性能对比与分析,实验结果表明BIRI是一种可行且高效的MSN路由机制.

The popularity of intelligent mobile terminals has greatly promoted the development of mobile social networks (MSNs). As the carrier of the terminal equipment, the human being has the feature of constantly moving which leads to dynamic changes of the network topology and brings many serious problems to MSN routing, such as high latency, low delivery rate and high overhead. In order to promote routing efficiency, based on the content-centric idea in information-centric networking (ICN) and the biogeography-based optimization (BBO) algorithm, an efficient BBO-inspired MSN routing algorithm with information-centric paradigm support (BIRI) is designed. Firstly, social metrics, social relationship strength and centrality, are redefined to direct BBO algorithm for community detection. Secondly, the novel strategies of content aggregation, data caching and bridge node selection are designed to support the efficient content retrieval and access. Based on these strategies, the enhanced intra-community and inter-community routing processes are proposed to release the interference caused by the mobility of the terminal equipment on data transmission. The proposed BIRI routing algorithm is simulated on the opportunistic network environment (ONE), and compared with other three baseline MSN routing algorithms and analyzed from three aspects of delivery rate, average latency and network overhead ratio. Experimental results show that the proposed BIRI mechanism is feasible and effective.