Relay Probability Characteristics in a Social Ad-Hoc Network with Different Intimacy Calculation Models and Social Network Structure Models

Abstract

We discuss a method of creating a wireless ad-hoc network in which nodes are connected by considering human relationships (i.e., intimacy) in social networks. Here, this is called a social ad-hoc network, and it could solve the incentive problem in which one has to relay the data of other people. Previous work showed that a social ad-hoc network is useful to connect to a specific destination through some nodes. A simple intimacy-based relationship model was evaluated with a criteria and relay probability. In order to better understand the relay probability characteristics of a social ad-hoc network, we use a variety of intimacy models in terms of intimacy distributions, indirect intimacy calculation methods and social network topologies. We then evaluate how the social ad-hoc network is useful. The results show a significant difference of 40% between different intimacy distribution models. We also discover that the density of the social network has a stronger effect on the relay probabilities than other parameters.