【摘要】 为了解决在消息敏感的机会网络中社团划分结果不可重用的问题,提出了一种与消息类型相匹配的机会网络分层模型.首先,将机会网络的物理节点集映射为与消息类型匹配的虚拟节点集,并以此为基础建立虚拟机会网络层;然后,在虚拟机会网络层上,建立虚拟节点集的社会关系;最后,对虚拟节点集的社会关系进行社团划分.实验结果表明:在消息数量相同的条件下,当消息序列中相邻位置消息的类型差异度分别为40%和100%时,在虚拟层上进行社团划分的时间与在物理机会网络上直接进行社团划分的时间相比分别减少约58%和89%;基于分层模型的社团划分的运行次数仅依赖于消息类型的数量,而不会随消息数量或消息序列中不同类型消息交错方式的变化而变化.更多还原

【Abstract】 In order to solve the problem that the community detection results cannot be repeatedly used in the information sensitivity opportunistic network,an opportunistic network hierarchical model which matches with information types is proposed. First,the physical node set in the opportunistic network is mapped as a virtual node set which matches with information types,based on which the virtual opportunistic network layer is established. Then,the social relationship of the virtual node set is built on the virtual opportunistic network layer. Finally,community detection is conducted on the social relationship of the virtual node set. The experimental results showthat when the difference degrees of the types of adjacent information in the information sequence are 40% and 100%,the time consumption in community detection on the virtual layer decreases by about 58% and 89% compared with that in the physical opportunistic network with the same information quantity,respectively. The execution number of the community detection operations based on the layer model only depends on the number of the information types,and does not change with the change of the information quantity or the overlapping ways of the different information types in the information sequence.更多还原