Network Quality of Service Through Local Adaptive Provisioning (LAP)
Scott C. Evans, T. Steven Markham, Ping Liu, Ishan P. Weerakoon · 2007
Maximization of End-to-End (E2E) Goodput in tactical networks requires coordination between edge, core, and configuration parameters. We propose an adaptive quality of service architecture that learns local Weighted Fair Queuing configuration settings based on queue depth and projected E2E success. Our adaptive strategy provides for local node optimization based on local node parameters and allows the network to flex to different settings as needed accommodating hot spots and traffic bottlenecks. The gain in E2E goodput for our adaptive learning method vs. global settings is quantified.