Passive Optical Networks (PONs) have been experiencing solid growth in the last few decades because were designed as an excellent alternative to solve the major problem for telecommunications networks: the bottleneck in access networks. The next generation of this technology, called Next-Generation PON (NG-PON), is the consequence of developments in optical technologies and provides support to the next generation services, improving TDMPONs
performance parameters and expanding the coverage area of these networks. This
geographical expansion benefits telecom companies that focus their efforts to simplify the network infrastructure by the unification of metropolitan, access and backhaul networks, decreasing the amount of local exchanges and even the number of nodes and, as a consequence, operational and financial costs. It means significant changes in the access networks scenario that tend to have large distances between ONUs and OLT, and a wide variety of services, making essential the presence of scheduling algorithms that are able to manage all shared resources efficiently, while ensuring fairness and the control in the upstream and downstream traffics dynamic allocation. It is through this context that this dissertation aims to present a hybrid grant scheduling algorithm based on priority queues, which gives efficiency and control to the OLT in the dynamic traffics scheduling. The results presented were obtained by the scenarios developed in a computational simulation environment and were based on the delay and throughput metrics for evaluating the algorithm
performance. It will be also evaluated how the amount of resources in the OLT interferes in these metrics.
|
