Original scientific paper

https://doi.org/10.17559/TV-20151109144645

Improved Algorithm for Distributed Points Positioning Using Uncertain Objects Clustering

Ivica Lukić orcid.org/0000-0001-7867-3385 ; Sveučilište J. J. Strossmayera u Osijeku, Fakultet elektrotehnike, računarstva i informacijskih tehnologija Osijek, Kneza Trpimira 2b, 31000 Osijek, Croatia
Mirko Köhler ; Sveučilište J. J. Strossmayera u Osijeku, Fakultet elektrotehnike, računarstva i informacijskih tehnologija Osijek, Kneza Trpimira 2b, 31000 Osijek, Croatia
Tomislav Galba ; Sveučilište J. J. Strossmayera u Osijeku, Fakultet elektrotehnike, računarstva i informacijskih tehnologija Osijek, Kneza Trpimira 2b, 31000 Osijek, Croatia

Abstract

Positioning of mobile objects that require communication with some kind of online service application is a very challenging task. Proper positioning with minimal deviation is an important mobile service system (MSS), e.g. taxi service used in this paper. It will perform all tasks for the users and reduce the overall travel distance. This paper is focused on the development of an algorithm that will find the optimal position for an MSS object and upgrade the system quality using uncertain data clustering. If the best position for the MSS is found, then the response time is short, and the system tasks could also be performed in usable time. The improved bisector pruning method is used for clustering stored data of mobile service system objects to provide the best position of system objects. As the best position of MSS objects, we use cluster centres. Using clustering, the total expected distance from end users to the service system is minimal. Therefore, the MSS is more efficient and has more time to fulfil additional tasks at the same time.

Keywords

clustering; data mining; Euclidian distance; information service; positioning

Hrčak ID:

219499

URI

https://hrcak.srce.hr/219499

Publication date:

24.4.2019.

Visits: 2.117 *