Analytical study of frame aggregation in error-prone channels

With many improvements in both the physical and MAC layers, the IEEE 802.11n standard aims to achieve a data transmission rate of up to 600 Mbps. The most important 802.11n MAC enhancement is frame aggregation which significantly reduces the headers overheads of the legacy MAC by aggregating multiple frames into a single large frame. In this paper we drive an analytical model to study the impact of the frame aggregation on the saturation throughput and access delay under error-prone channels. We describe the throughput as a function of physical data rate, and Frame Error Rate(FER). The analytical results show that the throughput and delay depend significantly on MPDU subframes size in high FER channels. Thus, We propose an algorithm which can dynamically adjust the MPDU subframe size based on the maximum FER tolerable by frame's access category. We use the feedback Block Acknowledgement (BACK) frame to compute the next MPDU subframe size which can takes into account the FER and the delay QoS requirements for the corresponding access category traffic.