A Kalman-fuzzy application for rate adaptation in IEEE 802.11 based multihop ad hoc networks

Rate adaptation has become one of the basic techniques in enhancing the capacity in todays multihop 802.11 WLANs. It is designed to cope with the status of wireless channels and achieve higher system spectral efficiency by exploiting the multi-rate capability provided by the IEEE 802.11 physical layer (PHY). As such, one of key aspects in designing rate adaptation scheme is the prior knowledge of the channel state. This becomes challenging when considering a fading channel conditions, where time varying channel gains cause random frame losses that need to be distinguished from those due to collisions and strong interference. In this paper, a new rate adaptation scheme that adopts a Kalman filter to predict the channel conditions is proposed. By predicting the future signal-to-interference noise ratio (SINR), the losses due to poor received signal and that of strong interference are taken into consideration. Furthermore, and to ensure correct operation of Kalman filter, the application of fuzzy logic controller has been explored for online tuning of Kalman Filter parameters. By tracking the covariance value of the measured interference instants, Q_t, the convergence of the Kalman filter can be observed and corrective measures using fuzzy logic can be applied to prevent divergence. The soundness of the proposed rate adaptation scheme is demonstrated by comparing with existing approaches through discrete event simulation environment.