Laboratory Study of Class-Based Queue Management Method on IP Network Routers

Abstract

It has been established that packet queue organization and servicing mechanisms are among the key technological tools for ensuring quality of service, traffic management, and allocation of link and buffer resources. However, a significant drawback of most existing queue servicing mechanisms is the considerable amount of administrative configuration required, which increases sharply as the number of organized queues grows. Therefore, automating the configuration of communication equipment using network programming tools in the Python environment is a viable solution to this problem. This approach enables the reduction of configuration time and the decrease in the likelihood of administrative errors. The study focuses on an optimization-based method for queue servicing. This method is based on optimizing flow-aggregation processes and on the balanced allocation of bandwidth among class-based queues. The article presents a methodology for a laboratory experiment. During the experiment, automated configuration functions were delegated to a network server running a Python environment. In the experiment, this server was responsible for collecting network-state information and calculating control parameters that determined the servicing order of class-based queues. As an example, the study implemented automated configuration for class-based queues created using the CBWFQ mechanism. The experimental results confirmed the adequacy of the computational model and the selected queue-servicing method. The server ensured the correct solution of optimization problems using appropriate Python libraries, and the resulting calculations were subsequently applied via network programming to remotely and automatically configure the corresponding router interface. Verification of the router state and its interfaces confirmed the correctness of the conducted experiment. In future work, the laboratory experiment methodology may be extended and supplemented with network testing tasks focused on analyzing current values of key packet quality-of-service indicators.