Radio resource allocation and hybrid multiplexing of voice and data over IP in aGSM/GPRS cellular network
Abstract In this study, a first order investigation of the issue of resource allocation between circuit voice, packet data and packet voice was completed. The study was done with reference to the GSM / GPRS air interface. To study the allocation of resources, suitable traffic source models were...
Main Author: | |
---|---|
Format: | Others |
Language: | en |
Published: |
2008
|
Online Access: | http://hdl.handle.net/10539/4787 |
id |
ndltd-netd.ac.za-oai-union.ndltd.org-wits-oai-wiredspace.wits.ac.za-10539-4787 |
---|---|
record_format |
oai_dc |
spelling |
ndltd-netd.ac.za-oai-union.ndltd.org-wits-oai-wiredspace.wits.ac.za-10539-47872019-05-11T03:41:54Z Radio resource allocation and hybrid multiplexing of voice and data over IP in aGSM/GPRS cellular network Vannucci, David Emmanuele Abstract In this study, a first order investigation of the issue of resource allocation between circuit voice, packet data and packet voice was completed. The study was done with reference to the GSM / GPRS air interface. To study the allocation of resources, suitable traffic source models were developed to represent the nature of the traffic offered to the base station subsystem. Circuit voice and packet data were represented using Markovian arrivals and exponentially distributed holding times. Voice over IP was modelled using a two-state Markov modulated Poisson process. The base station subsystem was modelled as a continuous time controller with eight channels (one GSM / GPRS TDMA frame). The radio propagation environment was considered by means of a large-scale propagation model, which would merely alter the load presented to the developed simulator package. From the results of the simulations, it was found that insufficient data resources lead to similar packet delay regardless of the packet size. It was found that if capacity on demand is used, then the data resources could equal the load. In the case of sufficient data channels, with capacity on demand, additional channels have a greater effect on average delay than the probability of it occurring. Prioritisation of VoIP packets did not significantly alter the probability of delay but affected the average packet delay. Packet size had a greater effect on average delay than the probability of delay. In the case of all eight channels being used for VoIP and data, the combined load should not exceed seven erlang, indicating that a higher voice load could be supported with VoIP than with circuit switched voice. 2008-05-07T12:49:36Z 2008-05-07T12:49:36Z 2008-05-07T12:49:36Z Thesis http://hdl.handle.net/10539/4787 en 1577657 bytes application/pdf application/pdf |
collection |
NDLTD |
language |
en |
format |
Others
|
sources |
NDLTD |
description |
Abstract
In this study, a first order investigation of the issue of resource allocation between
circuit voice, packet data and packet voice was completed. The study was done with
reference to the GSM / GPRS air interface. To study the allocation of resources,
suitable traffic source models were developed to represent the nature of the traffic
offered to the base station subsystem. Circuit voice and packet data were represented
using Markovian arrivals and exponentially distributed holding times. Voice over IP
was modelled using a two-state Markov modulated Poisson process. The base station
subsystem was modelled as a continuous time controller with eight channels (one
GSM / GPRS TDMA frame). The radio propagation environment was considered
by means of a large-scale propagation model, which would merely alter the load
presented to the developed simulator package. From the results of the simulations,
it was found that insufficient data resources lead to similar packet delay regardless
of the packet size. It was found that if capacity on demand is used, then the data
resources could equal the load. In the case of sufficient data channels, with capacity
on demand, additional channels have a greater effect on average delay than the
probability of it occurring. Prioritisation of VoIP packets did not significantly alter
the probability of delay but affected the average packet delay. Packet size had a
greater effect on average delay than the probability of delay. In the case of all eight
channels being used for VoIP and data, the combined load should not exceed seven
erlang, indicating that a higher voice load could be supported with VoIP than with
circuit switched voice. |
author |
Vannucci, David Emmanuele |
spellingShingle |
Vannucci, David Emmanuele Radio resource allocation and hybrid multiplexing of voice and data over IP in aGSM/GPRS cellular network |
author_facet |
Vannucci, David Emmanuele |
author_sort |
Vannucci, David Emmanuele |
title |
Radio resource allocation and hybrid multiplexing of voice and data over IP in aGSM/GPRS cellular network |
title_short |
Radio resource allocation and hybrid multiplexing of voice and data over IP in aGSM/GPRS cellular network |
title_full |
Radio resource allocation and hybrid multiplexing of voice and data over IP in aGSM/GPRS cellular network |
title_fullStr |
Radio resource allocation and hybrid multiplexing of voice and data over IP in aGSM/GPRS cellular network |
title_full_unstemmed |
Radio resource allocation and hybrid multiplexing of voice and data over IP in aGSM/GPRS cellular network |
title_sort |
radio resource allocation and hybrid multiplexing of voice and data over ip in agsm/gprs cellular network |
publishDate |
2008 |
url |
http://hdl.handle.net/10539/4787 |
work_keys_str_mv |
AT vannuccidavidemmanuele radioresourceallocationandhybridmultiplexingofvoiceanddataoveripinagsmgprscellularnetwork |
_version_ |
1719084868244078592 |