Delay tolerant networking in a shopping mall environment.
PhD thesis, University of Nottingham.
The increasing popularity of computing devices with short-range wireless offers new communication service opportunities. These devices are small and may be mobile or embedded in almost any type of object imaginable, including cars, tools, appliances, clothing and various consumer goods. The majority of them can store data and transmit it when a wireless, or wired, transmitting medium is available. The mobility of the individuals carrying such short-range wireless devices is important because varying distances creates connection opportunities and disconnections. It is likely that successful forwarding algorithms will be based, at least in part, on the patterns of mobility that are seen in real settings. For this reason, studying human mobility in different environments for extended periods of time is essential. Thus we need to use measurements from realistic settings to drive the development and evaluation of appropriate forwarding algorithms. Recently, several significant efforts have been made to collect data reflecting human mobility. However, these traces are from specific scenarios and their validity is difficult to generalize.
In this thesis we contribute to this effort by studying human mobility in shopping malls. We ran a field trial to collect real-world Bluetooth contact data from shop employees and clerks in a shopping mall over six days. This data will allow the informed design of forwarding policies and algorithms for such settings and scenarios, and determine the effects of users' mobility patterns on the prevalence of networking opportunities.
Using this data set we have analysed human mobility and interaction patterns in this shopping mall environment. We present evidence of distinct classes of mobility in this situation and characterize them in terms of power law coefficients which approximate inter-contact time distributions. These results are quite different from previous studies in other environments.
We have developed a software tool which implements a mobility model for "structured" scenarios such as shopping malls, trade fairs, music festivals, stadiums and museums. In this thesis we define as structured environment, a scenario having definite and highly organised structure, where people are organised by characteristic patterns of relationship and mobility. We analysed the contact traces collected on the field to guide the design of this mobility model. We show that our synthetic mobility model produces inter-contact time and contact duration distributions which approximate well to those of the real traces. Our scenario generator also implements several random mobility models.
We compared our Shopping Mall mobility model to three other random mobility models by comparing the performances of two benchmark delay tolerant routing protocols, Epidemic and Prophet, when simulated with movement traces from each model. Thus, we demonstrate that the choice of a mobility model is a significant consideration when designing and evaluating delay-tolerant mobile ad-hoc network protocols.
Finally, we have also conducted an initial study to evaluate the effect of delivering messages in shopping mall environments by exclusively forwarding them to customers or sellers, each of which has distinctive mobility patterns.
Thesis (University of Nottingham only)
||T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5101 Telecommunication
||UK Campuses > Faculty of Science > School of Computer Science
Hatton, Mrs Kirsty
||02 Feb 2015 09:20
||16 Sep 2016 01:52
Actions (Archive Staff Only)