Utility-Driven Traffic Engineering via Joint Routing and Rate Control for 802.11 MANETs
Gi Tae Kim, John R. Lee, Latha Kant, Alex Poylisher · 2021
Efficient network resource utilization together with sustained high Quality of Service (QoS) for applications is crucial in mobile ad hoc networks (MANETs) for both military and civilian deployments. In wired networks, traffic engineering mechanisms that jointly adapt routing and source rate control using primal-dual utility maximization have been shown to achieve optimal solutions [3], with distributed implementations. In this paper, we ask whether the same principled approach can work in CSMA (802.11) MANETs, propose the architecture and mechanisms that address unique challenges posed by shared spectrum and mobility, and show that the proposed approach outperforms the state of the art on several metrics in realistic scenarios.Our key contributions include: (a) an interference-aware multi-path selection algorithm based on the wireless channel bandwidth utilization and projected load, (b) a link cost function based on wireless channel occupancy that allows for rate adaptation to minimize both congestive and non-congestive loss, and (c) a rate adaptation mechanism robust to the loss of feedback packets. Our implementation is also the first to leverage the emergent P4-based network programmability for MANETs.