In a multi-rate IEEE 802.11 network, the conventional Distributed Coordination Function (DCF) Medium Access Control (MAC) aims to ensure max-min throughput fairness and equal channel access in scenarios with multiple nodes, while failing to satisfy air-time fairness. Consequently, nodes that have relatively poor channels or longer packets to transmit invade the channel substantially more than others, hence decreasing the throughput of nodes which have better channels or shorter packets. This phenomenon is known as the performance anomaly problem in the existing literature. In this thesis, we propose a novel distributed air-time fair algorithm to cope with the performance anomaly problem without having to change the conventional IEEE 802.11 DCF MAC. In the proposed algorithm, each node in the system runs multiple instances of the conventional IEEE 802.11 DCF back-off algorithm where the number of instances for the particular node is calculated independently from other nodes using only local information such as packet lengths and transmission rates. Both analytical and simulationbased results are provided to validate the effectiveness of the distributed air-time fair algorithm we propose.