Penggunaan Matrix Analytic Technique pada Perhitungan Parameter Kinerja Proses Handoff

Authors

  • R. Hadianti Departemen Matematika, FMIPA, Institut Teknologi Bandung. Jl. Ganesa 10, Bandung 40132, INDONESIA, Fax :062-22-2506450
  • J. Naiborhu Departemen Matematika, FMIPA, Institut Teknologi Bandung. Jl. Ganesa 10, Bandung 40132, INDONESIA, Fax :062-22-2506450
  • L. Dahliantini Departemen Matematika, FMIPA, Institut Teknologi Bandung. Jl. Ganesa 10, Bandung 40132, INDONESIA, Fax :062-22-2506450

DOI:

https://doi.org/10.5614/itbj.sci.2005.37.1.4

Abstract

This paper discusses performance evaluation of mobile cellular telecommunication system that is related to the handoff process. We focus our discussion to the system with fixed channel assignment and homogeneous traffic. Performance parameters we discuss in this paper are the blocking probability and the dropping probability of handoff signals, and the blocking probability of new originating call signals.
The operator of the system usually gives a service priority to handoff signals compare to new originating call signals. In this paper, the priority is given through channels reservation. A number of channels are reserved so that they can be used only for serving handoff signals. The remaining channels are used for serving new originating call signals and if there is an idle channel, it also can be used for serving handoff signals. With this priority, the arrival process and the service process of both types of signals can be modelled as a queueing network that consists of two queueing systems: one system for handoff signals and one system for both signals.
The performance of the queueing network is analyzed by deriving the a two-dimensional Markov process. The system of balance equations of the process, which is needed for obtaining the stationary probability distribution of the process, is a large system. We solve this system by applying Matrix Analytic Technique to the two sub-systems simultaneously. The performance parameter values, which are expressed by the stationary probability distribution, are obtained from the solution.

Downloads

Issue

Section

Articles