Abstract

With the advent of revolutionary wireless technologies and equipments, one may expect a great user mobility. In this paper, we study the mobility of users in user centric networking (UCN) environment. We propose the mobility prediction approach by WLAN group formation. We focus on mobility behavior of users and propose the WLAN group prediction approach. We base the proposed approach on the relatively simple parameters and evaluate the performance of our proposed concept by real implementation. Our approach is based on building the history of encountered WLAN groups, modeling an encounter instance of a new group as arrivals for the proposed model and finally adapting the prediction. In the granular level, the arrival rates are estimated using expectation maximization algorithm. We believe that the proposed approach is important for UCN like environments, where information on the hardware or software capabilities, available network resources, instant AP load and etc are not available for the end-user.

@INPROCEEDINGS{6554901,
author={Yildiz, M. and Khan, M.A.},
booktitle={Wireless Communications and Networking Conference (WCNC), 2013 IEEE},
title={Mobility behavior modeling in UCN},
year={2013},
pages={2184-2189},
keywords={mobility management (mobile radio);wireless LAN;WLAN;expectation maximization algorithm;mobility behavior modeling;mobility prediction;user centric networking;user mobility;wireless technologies;Hidden Markov models;Mathematical model;Mobile communication;Prediction algorithms;Predictive models;Stochastic processes;Wireless communication;AP Selection in UCN;Expectation Maximization Algorithm;Markov Arrival Processes;Mobility Behavior Modeling;Mobility Prediction},
doi={10.1109/WCNC.2013.6554901},
ISSN={1525-3511},}