Large effective call holding times for a PCS network

Yi-Bing Lin*, Imrich Chlamtac

*Corresponding author for this work

Research output: Contribution to conferencePaperpeer-review

11 Scopus citations

Abstract

This paper addresses two important PCS cost and performance issues: (1) Are blocking probabilities (for new call blocking, forced termination, or call completion) appropriate for PCS network performance measurement? (2) Should PCS service provider offer discount to phone calls that are forced terminated? In order to do that we derive the effective call holding time distributions for complete/incomplete calls by using an analytic model. We assume that the cell residence times have a general distribution so that the measured data from the field trials, after a second order approximation, can be used to drive our model. Our study on the effective call holding time distributions indicates that (1) blocking probabilities do not provide sufficient information to evaluate the performance of a PCS network, and that output measures, such as expected effective call holding times for complete/incomplete calls, are desirable; (2) for a PCS network with moderate user mobility, the service provider can offer substantial discounts to incomplete calls (to increase the user satisfaction) without significantly reducing the profit margin. A quantitative analysis is also provided that can be used to determine the discount factor.

Original languageEnglish
Pages143-147
Number of pages5
DOIs
StatePublished - 1 Dec 1996
EventProceedings of the 1996 7th IEEE International Symposium on Personal, Indoor and Mobile Radio Communications, PIMRC'96. Part 3 (of 3) - Taipei, Taiwan
Duration: 15 Oct 199618 Oct 1996

Conference

ConferenceProceedings of the 1996 7th IEEE International Symposium on Personal, Indoor and Mobile Radio Communications, PIMRC'96. Part 3 (of 3)
CityTaipei, Taiwan
Period15/10/9618/10/96

Fingerprint Dive into the research topics of 'Large effective call holding times for a PCS network'. Together they form a unique fingerprint.

Cite this