Distributed backpressure protocols with limited state feedback

The freedom and flexibility of wireless Mobile Adhoc Networks (MANETs) that make them extremely desirable for many military, emergency, and sensor network applications also present challenges for multiple layers in the network stack. Max-weight scheduling, also known as backpressure routing, is a cross-layer control algorithm that is well-known to be throughput optimal since it provides queue stability within the network for all traffic injection rates within the network's capacity. Despite its desirable properties like operating on instantaneous queue and channel states without requiring their statistics, max-weight scheduling relies on global knowledge of full system state information of the network. This is an overly optimistic assumption for most real-world implementations. In this work, we address this issue and develop a distributed max-weight scheduling algorithm in which nodes disseminate the necessary network state information for each to make the same optimal cross-layer control decision individually each time slot. We explore the idea of disseminating only a limited amount of state feedback for this algorithm and evaluate the subsequent impact on performance. We compare the distributed protocol to a centralized version that assumes nodes have global knowledge of this network state information, showing differences in performance from the increased overhead. We also introduce and evaluate an improvement to the algorithm that achieves better performance by dynamically adjusting values used to estimate queue backlogs in the limited state feedback scheduling algorithm.