Channel asymmetry and high fluctuation of channel conditions are two salient characteristics of wireless channels in mobile environments. Therefore, when using IEEE 802.11 devices in mobile environments, it is critical to have an effective rate adaptation scheme that can deal with these issues. In this paper, we propose a practical rate adaptation scheme called Rate Adaptation in Mobile environments (RAM) and implement it in the MadWifi device driver. RAM uses a receiver-based approach to handle channel asymmetry and a conservative SNR prediction algorithm to deal with high channel fluctuation. More importantly, RAM allows the receiver to convey the feedback information to the transmitter in a creative manner via ACK transmission rate variation, which does not require changes to the device firmware and hence is implementable at the device driver level. In addition, RAM adopts an effective scheme to guarantee that RAM-based and legacy IEEE 802.11 devices can interoperate with each other. The effectiveness of RAM is demonstrated via in-depth experimental evaluation in indoor static and mobile environments as well as outdoor vehicular environments.

Channel asymmetry and high fluctuation of channel conditions are two salient characteristics of wireless channels in mobile environments. Therefore, when using IEEE 802.11 devices in mobile environments such as vehicular networks, it is critical to have an effective rate adaptation scheme that can deal with these issues. In this paper, we propose a practical rate adaptation scheme called RAM (rate adaptation in mobile environments) and implement it in the Madwifi device driver. RAM uses a receiver-based approach to handle channel asymmetry and a conservative SNR prediction algorithm to deal with high channel fluctuation. More importantly, RAM allows the receiver to convey the feedback information in a creative manner via ACK transmission rate variation, which does not require changes to the device firmware and hence is implementable at the device driver level. The effectiveness of RAM is demonstrated through experimental evaluation in indoor static and mobile environments and outdoor vehicular environments, as well as simulation study based on SNR traces collected from the experiments.