Genius: um escalonamento baseado em algoritmos genéticos para comutadores de alto desempenho

Abstract

One of the most important components of network telecommunications is the router. Modern routers employ input-queued crossbar switches that require sophisticated scheduling techniques for packet transmission. The architecture of switches with input queues requires an scheduling process that establishes the transfer of packets from input to output ports. The performance of router depends of the scheduling algorithm, considering its throughput and complexity. In this work we survey the most relevant theoretical scheduling algorithms and propose a scheduling approach using genetic algorithms. We developed a simulator of discrete events for testing of schedulers performance. The proposed algorithm, called GENIUS, presents relevant performance and low complexity.