Algoritmo de escalonamento DRR com quantum adaptativo para o tráfego downlink de redes IEEE 802.16j

Detalhes bibliográficos
Ano de defesa: 2012
Autor(a) principal: Santos, Einar César
Orientador(a): Não Informado pela instituição
Banca de defesa: Não Informado pela instituição
Tipo de documento: Dissertação
Tipo de acesso: Acesso aberto
Idioma: por
Instituição de defesa: Universidade Federal de Uberlândia
BR
Programa de Pós-graduação em Engenharia Elétrica
Engenharias
UFU
Programa de Pós-Graduação: Não Informado pela instituição
Departamento: Não Informado pela instituição
País: Não Informado pela instituição
Palavras-chave em Português:
DRR
Link de acesso: https://repositorio.ufu.br/handle/123456789/14500
https://doi.org/10.14393/ufu.di.2012.343
Resumo: IEEE 802.16j networks provide, through the Relay Station (RS), improvements with regard to previous specifications of WiMAX in terms of increased coverage area, reduction of deployment costs due to low cost of RS compared to a Base Station (BS) and system average throughput increase. Its main purpose is to meet demands for wireless broadband access at lower cost. The efficient resource allocation is a challenge in the IEEE 802.16 standard and requires total commitment of scheduling algorithm. Few proposals for downlink scheduling developed for IEEE 802.16j networks are relevant to the present moment, and much of it disregards the maximum utilization of basic resources available, such as using only one RS due to a large traffic load, for example. In view of this, we propose the application of a Deficit Round Robin (DRR) algorithm with adaptive quantum operating together with a queue management and congestion control in RS. In the proposed DRR scheduling, implemented in the BS downlink scheduler, the quantum is calculated using the Maximum Transmission Unit (MTU) size and information about the congestion state in RS. In order to balance the average queue length in the output buffer of RS, we have implemented an algorithm based on Adaptive Random Early Detection (ARED). Finally, the proposed solution was evaluated using modeling and simulation. The simulation results demonstrate good performance of the proposed classes for UGS, rtPS, nrtPS and BE compared to a conventional DRR scheduling without queue management.