A comparison of queue management schemes in high performance Input-Queued switches

The Internet architecture provides a mechanism for protecting individual flows from congestion, by introducing queue management and scheduling algorithms in Input-Queued switches having finite queue size to improve the QoS performance in terms of throughput and delay. Queue management algorithms manage the length of packet queues by dropping packets when necessary or appropriate, while Scheduling algorithms determine which packet to send next and are used primarily to manage the allocation of bandwidth among flows. In this paper we consider two Queue Management methods: Drop Tail and Random Early Detection (RED) by scheduling with iSLIP scheduling algorithm. An important issue related to Input-Queued switch is scheduling cells to achieve high throughput as well as a low packet loss. We also analyze the QoS performance of these two schemes in terms of throughput, delay and packet loss. A simulation result shows that RED performs comparatively better than Drop Tail.