IMPROVING QoS OF VoWLAN VIA CROSS-LAYER BASED ADAPTIVE APPROACH

Voice over Internet Protocol (VoIP) is a technology that allows the transmission of
voice packets over Internet Protocol (IP). Recently, the integration of VoIP and
Wireless Local Area Network (WLAN), and known as Voice over WLAN
(VoWLAN), has become popular driven by the mobility requirements ofusers, as
well as by factor of its tangible cost effectiveness. However, WLAN network
architecture was primarily designed to support the transmission of data, and not for
voice traffic, which makes it lack ofproviding the stringent Quality ofService (QoS)
for VoIP applications. On the other hand, WLAN operates based on IEEE 802.11
standards that support Link Adaptive (LA) technique. However, LA leads to having a
network with multi-rate transmissions that causes network bandwidth variation, which
hence degrades the voice quality. Therefore, it is important to develop an algorithm
that would be able to overcome the negative effect of the multi-rate issue on VoIP
quality. Hence, the main goal ofthis research work is to develop an agent that utilizes
IP protocols by applying a Cross-Layering approach to eliminate the above-mentioned
negative effect. This could be expected from the interaction between Medium Access
Control (MAC) layer and Application layer, where the proposed agent adapts the
voice packet size at the Application layer according to the change of MAC
transmission data rate to avoid network congestion from happening. The agent also
monitors the quality of conversations from the periodically generated Real Time
Control Protocol (RTCP) reports. If voice quality degradation is detected, then the
agent performs further rate adaptation to improve the quality. The agent performance
has been evaluated by carrying out an extensive series ofsimulation using OPNET
Modeler. The obtained results of different performance parameters are presented,
comparing the performance ofVoWLAN that used the proposed agent to that ofthe
standard network without agent. The results ofall measured quality parameters have