基于节点中心性的移动机会网络网关选择算法

Gateway selection algorithms based on the centrality of nodes for mobile opportunistic networks

移动机会网络中,为使用户快速从因特网获取数据,需要选择某些合适节点担任网关。研究了移动机会网络网关选择问题。证明了k网关选择判定问题是NP难的。基于移动节点的中介中心性、度中心性、接近中心性,利用贪婪策略,提出了移动机会网络的网关选择算法。实验结果表明,基于节点中心性的3种网关选择算法均可改善网络性能。特别是基于节点接近中心性的算法可使性能得到最大幅度提升;基于节点中介中心性和度中心性的算法的性能受网络规模、需选择网关数目影响不大;基于节点接近中心性的网关选择算法对网络性能的改善幅度随需选择的网关数目的增多而增大。

In mobile opportunistic networks, suitable nodes are needed to be chosen to perform as gateways for users to obtain data from the Internet rapidly. The problem of gateway selection in mobile opportunistic networks is investigated. The deterministic problem of k gateway selection is proved to be NP hard. Based on the betweenness centrality, degree centrality, closeness centrality of mobile nodes, the gateway selection algorithms for mobile opportunistic networks are proposed by using greedy heuristic. Experiment results prove that, all the three gate selection algorithms can improve the network performance, especially the algorithm based on the closeness centrality of nodes can improve the performance most. The parameters of network size and the number of gateways do not influence the performance of the algorithms based on betweenness centrality and degree centrality much. As the number of gateways increases, the gateway selection algorithm based on the closeness centrality of nodes can improve the performance more.