Mobility Models for Next Generation Wireless Networks

Chapter 20

Social-Based Mobility Models

As seen in the previous chapter, human mobility—and the resulting pairwise contact patterns—is characterized by features that make it substantially different from mobility patterns generated by simple mobility models such as random walks. In particular, human mobility has been recently shown to be influenced by, and to influence, social interactions between individuals. Given the above observations, researchers have recently proposed mobility models specifically designed to take the interplay between human mobility and social interactions into account.

In this chapter, we will present a set of recently proposed social-based mobility models aimed at faithfully reproducing individual human mobility patterns, and the resulting pairwise contact patterns. We will start by presenting a weighted version of the popular random waypoint model aimed at modeling different degrees of popularity in waypoint selection, and a mobility model where node movement is biased toward sub-areas (communities) specific to each node. We will then introduce a community-based mobility model, where an input social network graph is used to shape the mobility pattern of nodes in a network. The fourth model considered builds upon the notion of “home location” resulting from recent studies on human mobility characterization, and proposes a tradeoff between distance from home and popularity of the destination to determine individual mobility patterns. Another model, inspired ...