The IEEE 802.11 MAC protocol specifies an enhanced distributed channel access (EDCA) mechanism with adjustable protocol parameters, providing differentiated access to different classes of wireless stations. The EDCA parameters are: CWmin and CWmax (minimum and maximum contention windows), AIFS (arbitration interframe space), and TXOP (transmit opportunity). This work studies the setting of these parameters in order to maximize capacity and optimize performance under QoS constraints of different classes.

We first revisit the analytical modeling in saturation and non-saturation conditions and provide more detailed calculations of mean channel access delay and throughput. Based on the analytical modeling, we examine service differentiation and the influence of EDCA parameters, for different load conditions. Then, considering two classes of wireless stations with higher and lower QoS demands, and optimizing with respect to average measures and constraints, we provide Pareto optimal pairs for the number of supported stations from the two classes, for different parameter set configurations. Further, we examine optimal parameter selection for a class of elastic traffic, in the presence of a delay-sensitive class whose parameters are fixed.

Our findings show that drastic service differentiation can undermine capacity of a WLAN. They also demonstrate that, in saturation conditions, TXOP is especially significant for maximizing throughput of an elastic traffic class, while AIFS for guaranteeing constraints of a delay-sensitive class. Conversely, in low-load conditions, the importance of these parameters is weakened and CWmin has the most central role in providing QoS guarantees. We summarize these findings as guidelines for the setting of 802.11e parameters in a scenario with data and real-time service classes.