Skip to Main content Skip to Navigation
Conference papers

Modeling scheduled data flow architecture in e-diagnosis networks: an open queuing network model approach

Abstract : Acquiring technical equipment high performance using e-maintenance technologies requires good control of data flow and high availability of dedicated networks such as VPN and associates. The problem considered in this paper applies to embedded multi-processor devices to improve e-monitoring, e-diagnostic, and e-prognostic of the components. Scheduled data flow architecture (SDF) executes instructions in the prescribed order in which they arrive, even though data might already be available for executing the instructions. In this paper, we propose to model the SDF using an Open queuing network with feedback. Two kinds of models are proposed. The first one has a single server, and the second one has multiple servers. A simulation will be performed using the “mean value analysis” method on the network of queues. The average response time will be computed for different synchronization and Execution units in the processor. The paper also compares the queuing model's utilization with multiple servers and the queuing model with a single server. The two models will depict the scheduling of instructions in the data flow architecture in an efficient way.
Complete list of metadata

https://hal.univ-lorraine.fr/hal-03091293
Contributor : Kondo Adjallah Connect in order to contact the contributor
Submitted on : Wednesday, December 30, 2020 - 7:30:56 PM
Last modification on : Tuesday, October 26, 2021 - 8:14:02 AM

Identifiers

  • HAL Id : hal-03091293, version 1

Collections

Citation

V. Bhaskar, Kondo Hloindo Adjallah. Modeling scheduled data flow architecture in e-diagnosis networks: an open queuing network model approach. IMS'2004 - International Conference on Intelligent Maintenance Systems: Advances in Maintenance and Modeling, Simulation, and Intelligent Monitoring of Degradation, Jul 2004, Arles, France. pp.CDROM, ISBN: 2-9522453-0-4. ⟨hal-03091293⟩

Share

Metrics

Record views

25