• Summary
• Related Documents
  • Active Bibliography
  • Co-citation
• Version History

Abstract:

As Asynchronous Transfer Mode (ATM) technology moves closer to deployment as a fundamental component of Broadband Integrated Services Digital Network (B-ISDN), attention is being focused on the technical issues related to providing customers with adequate Quality of Service (QOS) while using network resources efficiently.InPhase II of the Evaluation of Broadband Networking Technologies project, the University of Kansas (KU) conducted research that explored the relationships between service class definitions, Usage Parameter Control (UPC) or policing of customer traffic, network resource allocation mechanisms, and specific network conditions under which realistic QOS limits on cell loss ratio (CLR) can be met. As with Phase I of this study,performance evaluations were carried out using the Block Oriented Network Simulator (BONeS) tool. New policing, resource allocation and segmentation models were developed for Phase II, but a large number of BONeS models developed during Phase I were re-used, as planned. Acustomer access configuration was chosen to correspond with an early ATM deployment scenario. The four different traffic source models developed during Phase I (voice, video, data, image) were divided into two service classes: Reserved service, which included a guaranteed bandwidth allocation; and Standard service, which did not. An early-deployment traffic mix (dominated by LAN-LAN data traffic) was chosen along with a four-node network configuration and a three-hop connection pattern. The traffic was policed for compliance with agreed-upon characteristics in accordance with ATM Forum specifications. Reserved and Standard traffic streams were segregated using a dual-queue trunk service mechanism, which also gave cell loss priority to compliant Reserved traffic. Realistic QOS limits of 10

Citations