Research Article
MobiPLACE*: A Distributed Framework for Spatio-Temporal Data Streams Processing Utilizing Mobile Clients’ Processing Power
@INPROCEEDINGS{10.1007/978-3-319-11569-6_7, author={Victor Zakhary and Hicham Elmongui and Magdy Nagi}, title={MobiPLACE*: A Distributed Framework for Spatio-Temporal Data Streams Processing Utilizing Mobile Clients’ Processing Power}, proceedings={Mobile and Ubiquitous Systems: Computing, Networking, and Services. 10th International Conference, MOBIQUITOUS 2013, Tokyo, Japan, December 2-4, 2013, Revised Selected Papers}, proceedings_a={MOBIQUITOUS}, year={2014}, month={12}, keywords={}, doi={10.1007/978-3-319-11569-6_7} }- Victor Zakhary
Hicham Elmongui
Magdy Nagi
Year: 2014
MobiPLACE*: A Distributed Framework for Spatio-Temporal Data Streams Processing Utilizing Mobile Clients’ Processing Power
MOBIQUITOUS
Springer
DOI: 10.1007/978-3-319-11569-6_7
Abstract
The problem of continuous spatio-temporal queries’ processing was addressed by many papers. Some papers introduced solutions using single server architecture while others using distributed server one. In this paper, we introduce MobiPLACE, an extension to PLACE [13] system, a distributed framework for spatio-temporal data streams processing exploiting mobile clients’ processing power. We will extend the Query-Track-Participate (QTP) query processing model, introduced as a system architecture in PLACE, by moving the Query server role to mobile clients. This will reduce memory and processing load on our regional servers in exchange for a little additional communication and memory load on mobile devices. This makes the system more scalable and enhances average query response time. Improvements in mobile devices’ and communication links’ capabilities encouraged us to introduce this extension. In this paper, we will focus on range and k-NN continuous queries and their evaluation on MobiPLACE. Experimental study is made to compare between MobiPLACE* and PLACE* in terms of server response time and memory.