DQDB under heavy load: performance evaluation and fairness analysis
Marco Conti, Enrico Gregori, Luciano Lenzini
Abstract
Marco Conti, Enrico Gregori, Luciano Lenzini
Abstract
An extended simulation analysis of the asynchronous part of the distributed queue dual bus (DQDB) MAC protocol is reported. DQDB is analyzed in asymptotic conditions, i.e. bandwidth sharing among network nodes when each node is trying to seize all of the medium capacity. DQDB fairness and performance is investigated under a more realistic scenario. The results show that DQDB without the bandwidth balancing mechanism (BBM) is unfair and that the unfairness is dependent on the medium capacity and the bus length. When the BBM is enabled, DQDB fairness and performance, in the steady state, improve significantly. However, DQDB takes some time to achieve the steady state, during which time the node quality of service is degraded.>
OpenAlex reports 38 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
An extended simulation analysis of the asynchronous part of the distributed queue dual bus (DQDB) MAC protocol is reported. DQDB is analyzed in asymptotic conditions, i.e. bandwidth sharing among network nodes when each node is trying to seize all of the medium capacity. DQDB fairness and performance is investigated under a more realistic scenario. The results show that DQDB without the bandwidth balancing mechanism (BBM) is unfair and that the unfairness is dependent on the medium capacity and the bus length. When the BBM is enabled, DQDB fairness and performance, in the steady state, improve significantly. However, DQDB takes some time to achieve the steady state, during which time the node quality of service is degraded.>
Key concepts: Distributed-queue dual-bus, Computer network, Computer science, Asynchronous communication, Bandwidth (computing), Node (physics), Quality of service, Local area network