Control Theoretic Optimization of 802.11 WLANs: Implementation and Experimental Evaluation

Abstract

In 802.11 WLANs, the dynamic adaptation of the contention parameters along network conditions results in relevant performance improvements. Despite the ability to change these parameters has been available in standard devices for years, no adaptive mechanism using this functionality has been validated in a realistic deployment so far. In our work, we report our experiences related to the implementation and evaluation of two adaptive algorithms based on control theory, one centralized and one distributed, in a large-scale testbed consisting of eighteen commercial off-the-shelf devices. We conduct extensive measurements under non-ideal channel condition, considering the impact of different network scenarios in terms of number of active nodes and traffic generated. We show that both algorithms significantly outperform the standard configuration in terms of total throughput. We also expose the limitations inherent to distributed schemes, and demonstrate that the centralized approach substantially improves performance under a broad variety of scenarios, which confirms its suitability for real deployments.