Research On Scaling Law Of Wireless Social Networks

As the key part of network information theory,capacity analysis of ad hoc networks has become a hot topic among researchers since it provides guidance for network design and analysis of network performance. Different from that in point-point communication, capacity analysis of ad hoc networks is more challenging for its no center ?self-organization and dynamic topology. Kumar's work in capacity of wireless network is regarded as landmark in this area,it introduces the scaling law for analysis since the number of nodes in wireless networks tends to infinity. The scaling law of capacity in random networks is derived where nodes are uniformly distributed and select destination randomly. Scaling law of capacity in wireless social networks is main part of research on capacity of ad hoc networks. As the spreading of mobile devices such as smartphones and tablet PC, the extensive use of M2 M system, wireless social networks become combination of wireless communication and social networks. Services of wireless social networks rise and scale of them get larger, which makes them to be typical large-scale networks. The scaling law of capacity is a basic measuring standard for large-scale networks,thus the exploration of scaling law of capacity in wireless social networks is of great significance?Wireless social networks are the study object in this paper. Each individual in wireless social network not only acts as node in wireless communication but also plays a role in the social relationship. Therefore the transmission in wireless social network follows the rules in wireless communication while the selection of destinations has its tendency and follows some underlying laws,which is different from that the nodes choose their destinations randomly and equally in the existing study of network capacity. For ease of understanding the logic structure of wireless social networks, a more realistic model with three-layer architecture is built in this paper. It covers physical layer, social layer and session layer from bottom to top to make it convenient to analyze wireless social network. In physical layer, the transmission range of node and interference problem are fully considered. In social layer, the degree of node is set to be a variable accounting for the free scale property in social network. And the distribution of social contacts ate assumed to be based the geography location. In session layer, the frequency of the communication between source and its social contacts is set with regard to the distance between them.In this paper, the throughput capacity of wireless social network is derived based on the proposed model which takes the interaction of wireless communication and social networks into consideration. To derive the throughput capacity, a two-dimensional grid network is studied where and the probability of nodes' distribution can be converted to the area of the square. Results show that geography and clustering property have impact on the throughput capacity of wireless social network.