Medium access control (MAC) in existing wireless LAN systems exercises an upper bound on the packet size known as Maximum Transmission Unit (MTU). Packet fragmentation has been proposed in the IEEE 802.11 standard to counter errors stemming from channel impairments and collisions with other users. Though smaller packet sizes reduce the probability of packet loss, they also simultaneously increase the overhead associated with sending the data. Further, existing MAC standards provide little or no quality of service (QoS) to video applications, particularly in terms of catering to the inconsistent distortion contributed by the video packets. In this thesis, we evaluate the performance of video priority-aware packet size allocation and transmission using IEEE 802.11e Hybrid coordination function Controlled Channel Access (HCCA). Our scheme trades off the packet fragment sizes with the available overhead; smaller fragment sizes (more overhead) for highest priority and larger fragment sizes (less overhead) for lowest priority packets. We perform simulation in ns-2 and evaluate the goodput for each priority and compare it to the case (i.e. non-QoS) where a single packet size is used for all video priority packets. We also plot the received video quality in terms of PSNR for different channel BER and different number of video users. Our priority-aware scheme provides minimum of 0.5dB to maximum of 2dB gain over the optimal packet size scheme and provides minimum gain of 2dB to maximum gain of 9dB gain over baseline scheme which uses MTU size for transmission
