Service-based rate adaptation architecture for IEEE 802.11e QoS networks

Abstract

One of the promising solutions to combat wireless channel fluctuations in IEEE 802.11 networks is to appropriately exploit multi-rate capability of the physical layer. The mechanism to adaptively select actual transmission rates based on channel condition is generally referred to as RA (rate adaptation). Recently, T-RA (throughput-based rate adaptation) has been proposed to maximize system capacity. On the other hand, for multimedia traffic, E-RA (error-based rate adaptation) has been proposed, which guarantees target error rate. In this paper, we newly propose S-RA (service-based rate adaptation) which dynamically selects an RA method suitable for the traffic type of transmission packets. T-RA is employed for best-effort data while E-RA for real-time multimedia. Thus, in S-RA, different RA schemes are adaptively chosen on per-packet basis. The proposed S-RA architecture can be incorporated into the 802.11e QoS MAC where traffics are classified into multiple classes by their QoS requirements. Combined with 802.11e, the S-RA architecture can contribute to better QoS enhancement of real-time traffic, while preserving consistent BE throughput when compared with only T-RA or E-RA