Abstract:
Wireless VoIP is becoming an increasingly important application in recent
years. This fact, coupled with the increasing interest in location based ser-
vices, strongly suggest that tracking of wireless VoIP clients will become a
widely deployed feature in emerging wireless applications. In this paper, we
evaluate the performance of voice and tracking sessions analytically and via
simulations for the scenario to especially support VoIP and tracking applica-
tions. We carry out experiments to further prove the validity and strengthen
the proposed analytical and simulation models. We rst determine an up-
per bound on a maximum number of wireless tracking clients that can be
maintained under a single access point. We change the transmission fre-
quency of tracking information and examine that it has a signi cant impact
on the tracking capacity. We further extend this study to evaluate the ca-
pacity of location tracking of wireless VoIP clients, to investigate the e ect of
transmission frequency over the capacity of combined VoIP and tracking ses-
sions. From various performance measurements; we develop an insight that
at higher packetization intervals (e.g. 60ms) of VoIP tra c, the capacity
of combined VoIP and tracking sessions decreases by 30% compared to the
VoIP only capacity. The presented study further demonstrates that at lower
transmission frequency (e.g. 500ms) of tracking information, the capacity of
combined VoIP and tracking sessions coincides to the VoIP only capacity.
Finally, we evaluate the capacity of the VoIP and tracking wireless clients
using UDP (User Datagram Protocol) and DCCP (Datagram Congestion
Control Protocol) in the presence of TCP (Transmission Control Protocol)
tra c. Our studies propose that compared to UDP, DCCP not only im-
proves the combined VoIP and tracking capacity but also enables TCP to
get a reasonable bandwidth share.
