On the performance of rate control algorithm Minstrel

IEEE 802.11 wireless systems usually contain rate control algorithms which are designed to adapt the transmission rate between several available rates in response to varying channel conditions. The efficiency of the rate control algorithm in selecting optimal data rates for the channel conditions directly impacts on the throughput of the wireless system. Minstrel is a rate control algorithm that has good performance compared with other algorithms, and is widely implemented in popular wireless drivers such as MadWiFi, Ath5k and Ath9k. However, surprisingly, there is very little literature studying the performance of Minstrel. In this paper, we present an experimental study using a commercial wireless access point which evaluates the performance of Minstrel against fixed rates in a real-world IEEE 802.11g testbed. This experimental study takes a number of factors into account, such as uplink and downlink channels, transmission power, channel dynamics, Access Point (AP) location, and datagram size. The results show that while Minstrel performs well in many cases (particularly with “good” or “stable” channel conditions), the algorithm has difficulty achieving optimal throughput performance with “poor” or highly “dynamic” channel conditions. Our findings provide useful information on the design of rate control algorithms.