The Research On Hierarchical Architecture Of Opportunistic Networks Based On Community Detection

Opportunistic network derives from DTN network, which means communication in a non-fully connected network environment can be ensured. Data transmission in opportunistic network relies on chance encounters between mobile nodes. The main idea of the data transmission is "store-carry-forward". The main feature is the discontinuity of the network topology, which can meet the real requirements of the self-organizing networks better. The architecture of the opportunistic network has a great impact on network performance. According to the characteristics of the network, identifying the network structure to establish reasonable network architecture will directly affect the performance of the network, and will impact the design of routing as well. Meanwhile, the architecture will provide a guideline to resolve the scalability of the opportunistic network. However, the innovation of existing opportunistic network architecture is rare. Handheld devices carried by people forms an opportunistic network which maps the relationship of the human society. Understanding the structure of the network of community helps to optimize network performance, thus improving the efficiency of the data transmission in the network. In this paper, the layered architecture of opportunistic network is built based on the network structure identification, which has great theoretical significance and application value.First, static cluster nodes are set on hot spots in advance and an improved distributed community detection algorithm SIMPLE-S is proposed in this paper to identify geographic communities. The algorithm of community detection is achieved in the perspective of each static cluster nodes, and both the link aging between nodes and dynamically change coursed by outdated nodes are considered. The algorithm also considered the relationship community and geographic location. The simulation results show that the algorithm can effectively identify dynamic geographic communities in the opportunistic network.Second, a hierarchical architecture of opportunistic network based on community detection is proposed which is achieved by encoding the geographic communities and static cluster nodes, considering the community relations, location and time.Finally, a geographic community routing (GCR) is proposed in this paper based on the hierarchical architecture of opportunistic network. The basic principle is to take advantage of encoding geographic community information in the hierarchical architecture of opportunistic network to select relay nodes, considering community relations, time and location. Simulation results show that the geographic community routing strategy can effectively improve the success rate of data transmission and reduce the delay time of data transmission. The effectiveness of the routing strategy also verified that the hierarchical architecture of opportunistic network is able to effectively improve network data transmission ability based on identified communities.