A dynamic CBWFQ scheme for service differentiation in WLANs

In this paper we address the issue of service differentiation in wireless networks based on the IEEE 802.11 standard. Wireless local area networks (WLANs) of this type are becoming increasingly prevalent, counting millions of deployed networks in homes, offices and hot spots. However, the medium access control (MAC) protocol and the unpredictability of the wireless channel prohibit the direct employment of classical wireline service differentiation algorithms. The main contribution of this paper is a network layer mechanism, capable of differentiating service, while improving fairness and aggregate throughput in the presence of both uplink (TCP) and downlink traffic. We exploit a class based weight fair queuing (CBWFQ) scheme, properly supplemented by two algorithms that perform periodical weight adjustments. The two algorithms deal effectively with important WLANs problems, such as, unfairness due to location dependent channel errors, uplink-downlink unfairness and decreased throughput for all nodes, when a station transmits at a small rate. The proposed solution has been implemented in a testbed and the conducted experiments confirmed its resultfulness in real life scenarios