CAPACITY OF PRESENT NARROW-BAND CELLULAR-SYSTEMS - INTERFERENCE-LIMITED OR BLOCKING-LIMITED

Citation
Vm. Jovanovic et J. Gazzola, CAPACITY OF PRESENT NARROW-BAND CELLULAR-SYSTEMS - INTERFERENCE-LIMITED OR BLOCKING-LIMITED, IEEE personal communications, 4(6), 1997, pp. 42-51
Citations number
12
ISSN journal
10709916
Volume
4
Issue
6
Year of publication
1997
Pages
42 - 51
Database
ISI
SICI code
1070-9916(1997)4:6<42:COPNC->2.0.ZU;2-F
Abstract
There is a widespread notion within the North American cellular indust ry that present cellular systems are intrinsically interference-limite d. This idea in part comes from a number of reported simulations which were based on the assumption that mobiles always connect to the neare st base station. In this article we present analytical and simulation results for the case where mobiles are connected to the best base stat ion (the one with the strongest signal). At the 10 percent outage leve l, it is shown that the nearest server scenario tends to underestimate the carrier-to-interference (C/I) ratio by as much as 5 dB (with resp ect to the best server case). Some system imperfections which are not traditionally assessed by such simulations are further discussed. In p articular, nonideal positioning of the base stations was found to have negligible impact on C/I if the traditional cellular rule-of-thumb cr iteria are obeyed. Nonuniform spatial distribution of the mobiles was analyzed based on real data from the Toronto area. It was found to hav e negligible impact on the average C/I performance. Blocking problems in such a case, however, could easily reduce the theoretical capacity with uniform loading by as much as 30 percent. Finally, C/I deteriorat ions due to nonideal handoffs are examined, This analysis is very sens itive to the handoff mechanism implementation details. Although the re sults for hand-offs should be regarded with some caution, in view of a ll the results presented in this article we feel that cellular systems are in general limited by blocking rather than by interference.