SciELO - Scientific Electronic Library Online

vol.8 issue3Quality Control of Ceramic Pieces by Means of Vibration Analysis and Artificial Neuronal Networks author indexsubject indexsearch form
Home Pagealphabetic serial listing  

Services on Demand




Related links

  • Have no similar articlesSimilars in SciELO


Computación y Sistemas

Print version ISSN 1405-5546


CASTANEDA TRUJILLO, Omar  and  GARCIA HERNANDEZ, Carlos Felipe. Neural Network Based Wireless ATM Scheduler for Predicting and Shaping VBR and ABR Traffics. Comp. y Sist. [online]. 2005, vol.8, n.3, pp.176-186. ISSN 1405-5546.

Managing of the bandwidth resources in a WATM system, it results more critical than in an ATM system, due to the resources scarcity and to the transmission medium with lesser reliability. Therefore, traffic and congestion control mechanisms, more secure than the ones used in ATM, are required. This research work is aimed at the development of a wireless ATM scheduler based on the traffic prediction, using artificial neural networks at the ATM switch entrance, in order to control and shape the traffic in the switch exit connection (wireless link). Consequently, this reduces the congestion conditions mainly originated for the bursty traffic from the real time and non-real time ABR and VBR services. Connection Admission Control and ABR Flow Control mechanisms and applying delays in cells to manage the switch exit queue's buffer occupancy, have been considered, in order to control and shape the output traffic, reducing the congestion conditions and improving the wireless-link bandwidth utilization distribution between the VBR and ABR services.

Keywords : ATM, wireless; WATM; scheduler; artificial neural network; neuroscheduler; traffic prediction; traffic shaping; real time traffic; VBR; ABR; flow control; traffic control; congestion control; CAC; connection admission control; Hydragyrum; simulation; quality of service; QoS and buffer occupancy.

        · abstract in Spanish     · text in Spanish     · Spanish ( pdf )


Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License