Performance optimization in IEEE 802.11 WLANs with co-channel interference

IEEE 802.11 wireless local area networks (WLANs) are increasingly invoked to provide broadband internet access to mobile and wireless devices. Today, the ubiquity of WLANs is such that WLAN access points (APs) are deployed densely, especially in indoor environments; so much so that often several APs operating in close geographical proximity must contend for channel bandwidth due to the limited number of orthogonal channels available for WLANs. This so-called co-channel interference, nowadays inevitable in WLANs, causes severe throughput degradation, and mitigating this problem has become an important step towards improving the network performance. In this thesis, motivated by real world experiments and simulation results, we study how co-channel interference impacts the network throughput, and propose solutions to alleviate the induce problems.;To begin with, we model the MAC layer activities of a single-cell WLAN under the influence of hidden terminals, and identify the causes of unfair bandwidth allocation. Unlike existing models, our model can accommodate different numbers of hidden nodes without increasing the model complexity. Given any number of hidden nodes, only four constraints are needed to describe the interaction between stations and the AP with the consideration of both uplink and downlink traffic. Based on our model, we formulate a bandwidth allocation problem to optimize the network throughput and fairness under some predefined requirements by systematically tuning the AP's and stations' contention windows.;Then, we extend this study to multi-cell WLANs, demonstrating via simulation and real world test-bed experiments, that a severe throughput imbalance occurs between downlink TCP flows even in the simplest of multi-cell WLANs where only two mutually hidden APs compete for channel access. To solve this unfairness problem, we derive an analytical model that describes the interaction between TCP flows at the MAC layer, and formulate the throughput allocation problem as a nonlinear optimization problem subject to certain fairness requirements. Our formulation considers real world complexity such as hidden terminals, packet transmission retry limit, and the unique characteristics of TCP traffic. Solving our optimization problem yields the optimal MAC layer contention window settings that can lead each TCP flow to its target end-to-end throughput without any per-flow queuing or modification of the TCP sender.;Finally, we generalize our study to multi-cell WLANs with arbitrary topologies in the presence of the hidden terminals and spatial unfairness. Combining the rate control and contention resolution, we formulate the optimal rate allocation problem atop the CSMA/CA protocol as a non-convex optimization problem. Unlike previous approaches that require maximal weight scheduling or ignore the hidden terminal problem, our formulation considers a realistic IEEE 802.11-based MAC layer model including random backoff, carrier sensing, frame retransmission and contention window (CW) setting. We propose a simple scheme to transform this non-convex problem into a convex one, and derive a distributed algorithm to obtain the maximum transmission rate and the optimal contention window setting. To further improve the network performance, we incorporate association control into the cross-layer optimization problem, reducing the number of hidden terminals by discarding problematic associations and optimally re-associating APs and clients. We formulate such a problem as a non-convex mixed integer programming problem, which is known to be NP-hard and propose a distributed algorithm to approximate the optimal solution of this problem.