Using social network theory for modeling human mobility

Self-Organized Mobile Communication Technologies and Techniques for Network Optimization

The 5th Generation (5G) of mobile communication networks is being developed to address the demands and business contexts of 2020 and beyond. Its vision is to enable a fully mobile and connected society and also to trigger socio-economic transformations in ways eventually unimagined today. This means that the physical world to be covered with planned 5G networks including communication networks, humans and objects is becoming a type of an information system. So as to improve the experience of individuals, communities, societies, etc. within such systems, a thorough comprehension of intelligence processes responsible of generating, handling and controlling those data is fundamental. One of the major aspects in this context and also the focus of this chapter is the development of novel methods to model human mobility patterns, which have crucial role in forthcoming communication technologies. Human mobility patterns models can be categorized into synthetic, trace-based and community-ba...

2011

In this paper we introduce Spatio-TEmporal Parametric Stepping (STEPS) - a simple parametric mobility model which can cover a large spectrum of human mobility patterns. STEPS makes abstraction of spatio-temporal preferences in human mobility by using a power law to rule the nodes movement. Nodes in STEPS have preferential attachment to favorite locations where they spend most of their time. Via simulations, we show that STEPS is able, not only to express the peer to peer properties such as inter-contact/contact time and to reflect accurately realistic routing performance, but also to express the structural properties of the underlying interaction graph such as small-world phenomenon. Moreover, STEPS is easy to implement, flexible to configure and also theoretically tractable.

Mobility is attracting more and more interests due to its importance for data forwarding mechanisms in many networks such as mobile opportunistic network. In everyday life mobile nodes are often carried by human. Thus, mobile nodes’ mobility pattern is inevitable affected by human social character. This paper presents a novel mobility model (HNGM) which combines social character and Gauss-Markov process together. The performance analysis on this mobility model is given and one famous and widely used mobility model (RWP) is chosen to make comparison..

2012

Abstract: We propose to use social networking data to validate mobility models for pervasive mobile ad-hoc networks (MANETs) and delay tolerant networks (DTNs). The Random Waypoint (RWP) and Erdos-Renyi (ER) models have been a popular choice among researchers for generating mobility traces of nodes and relationships between them. Not only RWP and ER are useful in evaluating networking protocols in a simulation environment, but they are also used for theoretical analysis of such dynamic networks.

Journal of Network and Computer Applications, 2014

Opportunistic networks (OppNets) are modern types of intermittently connected networks in which mobile users communicate with each other via their short-range devices to share data among interested observers. In this setting, humans are the main carriers of mobile devices. As such, this mobility can be exploited by retrieving inherent user habits, interests, and social features for the simulation and evaluation of various scenarios. Several research challenges concerning human mobility in OppNets have been explored in the literature recently. In this paper, we present a thorough survey of human mobility issues in three main groups (1) mobility characteristics, (2) mobility models and traces, and (3) mobility prediction techniques. Firstly, spatial, temporal, and connectivity properties of human motion are explored. Secondly, real mobility traces which have been captured using Bluetooth/Wi-Fi technologies or location-based social networks are summarized. Furthermore, simulation-based mobility models are categorized and state-of-the art articles in each category are highlighted. Thirdly, new human mobility prediction techniques which aim to forecast the three aspects of human mobility, i.e.; users' next walks, stay duration and contact opportunities are studied comparatively. To conclude, some major open issues are outlined.

Wireless Personal Communications, 2011

A Personal Network (PN) is a user-centric design interconnecting numerous devices belonging to a user in different geographic locations, such as home, office, car, etc., to form a single global network to offer fully personalized services. In PNs devices of the user move in different groups, where these groups merge and split. In this paper, we design and simulate a PN Mobility Model (PNMM) to capture the characteristics of movements of devices in PNs. We propose a simple stability evaluation method for group mobility models and apply the method to PNMM. Through the stability evaluation, we find that PNMM possesses long-term steady state behavior. Moreover, for the evaluation of mobility models, some evaluation methods have been proposed, which include non-homogenous node mobility, relative node mobility in a group, and dynamics of group merging and splitting. Analysis shows that PNMM is better than other models to represent the PN mobility properties. In addition, the impact of PNMM on the performance of a PN Clustering Protocol (PNCP) has been studied. Simulation results provide insights into the performance of PNCP, and provide guidelines for future design of PN clustering. PNMM can be easily applied to any Personal Area Network or Body Area Network with slight modifications.

2007

We conduct a statistical study of human mobility using over 1000 hours of GPS traces of human walks involving 44 participants in five different locations, two university campuses, state fair, New York City, and Disney World. Our data reveals statistical features similar to those in what physicists have long called Levy random walks (or Levy walks). These features include heavy-tail distributions of flight lengths and super-diffusive nature of mobility. We find that these tendencies are likely caused by human intentions in deciding travel destinations (and distance and sojourn time thereof), but not by geographical constraints such as roads, buildings, boundaries, etc, and that geographical constraints, instead, cause truncations of flight lengths and discontinuity in the statistical tendency of mobility statistics. Based on these findings, we construct a simple Levy walk mobility model that emulates human walk patterns expected in outdoor mobile networks carried by humans. Based on these findings, we construct a simple Levy walk mobility model that emulates human walk patterns expected in outdoor mobile network environments. We demonstrate that the Levy walk model can be used to recreate the statistical patterns commonly observed in previous mobility studies such as the power-law distributions of human inter-contact times and that the simulation performance of mobile network routing protocols under the Levy walk model exhibits distinctive performance features unexplored under existing mobility models.

We want to understand the impact of social relationships on mobility models that capture the essence of patterns in human movements. In this thesis, we develop data-driven models that demonstrate the effect of friendship on human mobility and capture how do humans move as a function of social ties. To model and analyze human mobility, we propose to use digital footprints left and shared on social networking media, analyze individual footprints to provide observations on how geoproximity impacts friendship and correlate to network communities, and use social relationships to model individual and population mobility. For an application, we apply our work in human mobility to supplement the state of the art research for mobile networks by simulating different mobility models to understand the effect of movements on network congestion. v

2011

This paper goes over work under development in the context of mobility modeling, Internet architectures and services. We express an opinion concerning aspects related to an adequate modeling of human movement and to the relevancy of such modeling in the overall Internet operation. We give insight into the most recent trends on social mobility modeling and discuss the impact that such modeling has/may have be it on technology, or on society.

2013 Proceedings IEEE INFOCOM, 2013

Mobility is one of the main factors affecting the design and performance of wireless networks. Mobility modeling has been an active field for the past decade, mostly focusing on matching a specific mobility or encounter metric with little focus on matching protocol performance. Recently, new generations of wireless and mobile services have emerged based on communal aspects of mobile societies; conventionally not considered in mobility modeling. This study investigates the adequacy of existing mobility models in capturing various aspects of human mobility behavior (including communal behavior), as well as network protocol performance. This is achieved systematically through the introduction of a framework that includes a multi-dimensional mobility metric space to measure individual, pair-wise (encounter) and group (community) metrics.