Simulations framework for network congestion avoidance algorithms using the OMNeT++ IDE

This paper presents a new approach for developing network congestion avoidance simulations using the OMNeT++ IDE. Two active queue management and congestion avoidance algorithms are proposed for simulation, they are the WRED (Weighted Random Early Detection) and the improved WRED algorithms. In the case of the improved WRED algorithm, the QoS-based queues´ lengths are dynamically adjusted to accommodate the current data flow. Both algorithms were tested in the OMNeT++ IDE and the results of the simulations were compared to the ones obtained previously using a physical test network. They showed that the improved WRED algorithm has increased performance over the simple WRED algorithm. Also, by obtaining similar results with the physical implementation, OMNeT++ proves to be a reliable environment for developing frameworks for network congestion avoidance algorithms.