Since the ascendance of CDMA (Code Division Multiple Access) as the generally accepted
multiple access technology of choice for next generation networks, the rate of development of
the wireless cellular communication industry has been phenomenal. Next generation wireless
cellular networks offer a myriad of voice, video, data and text based information services for the
future multimedia and information society. This mixed service scenario implies that the same
[mite resource i.e. the air interface must be shared amongst different classes of user, each with a
specific quality of service. With multiple services competing for the same resource and with the
inherent soft capacity nature of CDMA, call admission control becomes a formidable task. The
problem is further compounded by the introduction of priorities between classes. Call admission
control is an essential component of these next generation networks and the open nature of the
current standards, such as UMTS (Universal Mobile Telecommunication System) allow for
vendor implementation of different call admission control policies.
The main area of focus in this dissertation is on a proposed downlink, load-based, mixed service
call admission policy. In a CDMA environment with symmetrical service and equal bandwidths
in each direction, the uplink is commonly considered to be the bottleneck. Based on the
asymmetric nature of the expected traffic in next generation networks the downlink is envisaged
as the future bottleneck. Some of the more common choices for downlink call admission control
include number based as well as power based call admission policies. A load-based call
admission policy has been chosen as the maximum load threshold that can be supported varies
with the state of the system and thus effectively models the behaviour of a soft capacity CDMA
network. This dissertation presents a teletraffic performance analysis model of a load-based call
admission control policy for downlink mixed service CDMA cellular networks.
The performance analysis yields customer oriented grade-of-service parameters such as call
blocking probability which is essential for network planning. In our analysis we incorporate a
Birth-Death Markov queuing model. This mathematical model is verified though computer
simulation. / Thesis (M.Sc.Eng.)-University of KwaZulu-Natal, 2007.
Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:ukzn/oai:http://researchspace.ukzn.ac.za:10413/8990 |
Date | January 2007 |
Creators | Ramlakhan, Niven Bhimraj. |
Contributors | Takawira, Fambirai. |
Source Sets | South African National ETD Portal |
Language | en_ZA |
Detected Language | English |
Type | Thesis |
Page generated in 0.0014 seconds