kim-jclassify - ORIGINAL ARTICLE Periodic properties of...

Info iconThis preview shows pages 1–2. Sign up to view the full content.

View Full Document Right Arrow Icon

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: ORIGINAL ARTICLE Periodic properties of user mobility and access-point popularity Minkyong Kim David Kotz Received: 11 August 2005 / Accepted: 16 November 2005 Springer-Verlag London Limited 2006 Abstract Understanding user mobility and its effect on access points (APs) is important in designing loca- tion-aware systems and wireless networks. Although various studies of wireless networks have provided useful insights, it is hard to apply them to other situa- tions. Here we present a general methodology for extracting mobility information from wireless network traces, and for classifying mobile users and APs. We used the Fourier transform to reveal important periods and chose the two strongest periods to serve as parameters to a classification system based on Bayes theory. Analysis of 1-month traces shows that while a daily pattern is common among both users and APs, a weekly pattern is common only for APs. Analysis of 1-year traces revealed that both user mobility and AP popularity depend on the academic calendar. By plotting the classes of APs on our campus map, we discovered that their periodic behavior depends on their proximity to other APs. Keywords Wireless network User mobility Popularity of access points Periodicity 1 Introduction Wireless networks have become popular and are get- ting more attention as a way to provide constant con- nectivity over a large area in cities and as an inexpensive way to provide connectivity to rural areas. The growing popularity of wireless networks encour- ages the development of new applications, including those that require quality of service (QoS) guarantees. To provide QoS, it is often useful to predict user mobility. We also need simulators of wireless network environments to test these new applications and these simulators require user mobility models. Thus, we aim to understand mobility of mobile devices in WiFi networks. As more mature wireless networks become avail- able, several studies of wireless networks have been published, including studies of a campus [ 7 , 8 , 11 ], a corporate environment, and a metropolitan area. Henderson et al. [ 7 ] analyzed the characteristics of wireless network usage on the Dartmouth campus using traces collected during the Fall 2003 and Winter 2004 terms. Balazinska and Castro [ 2 ] traced 1,366 corporate users on 117 APs over 4 weeks. Tang and Baker [ 12 ] studied a 7-week trace of the Metricom metropolitan-area packet radio wireless network, containing 24,773 mobile radios. Although these stud- ies help us to understand characteristics of different network environments and user groups, it is often dif- ficult to apply the findings of these studies to other applications. So we set out to develop methods to ex- tract mobility characteristics from network traces, allowing anyone to obtain model parameters from traces of their network (or a network similar to the desired network)....
View Full Document

This note was uploaded on 05/27/2011 for the course CIS 4930 taught by Professor Staff during the Spring '08 term at University of Florida.

Page1 / 15

kim-jclassify - ORIGINAL ARTICLE Periodic properties of...

This preview shows document pages 1 - 2. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online