Juan Rendon, Thomas Plückebaum, Stephan Jay
LRIC Cost Approaches for Differentiated QoS in Broadband Networks
No. 325 / Dezember 2009
Summary
The goal of this research project is to discuss how the utilization of Quality of Service (QoS) techniques should be included in an LRIC cost model. Several operators are deploying or have deployed QoS techniques that help improve qualified service production and that can have an impact on the cost of the service. On the one hand, one of the key points that should be addressed in the elaboration of an LRIC cost model is the definition of an efficient network. On the other hand, a question that should be studied is how QoS concepts should be incorporated into an LRIC cost model. To answer these questions the report has been divided into three sections
The principles used for the design of an efficient network are described in the first part of the report. Network design, network engineering and traffic engineering tools can be used for the design of a network with QoS capabilities. The main QoS solutions described are PSTN QoS solutions, Layer-2 QoS solutions, IP QoS solutions, and transport-layer and application-layer solutions. Network operators implement several solutions concurrently. The report discusses also the soft-assurance and hard-assurance models for QoS provisioning. Both QoS models are differentiated by the service that can be provided during normal and abnormal network conditions and it is expected that the cost of deploying QoS mechanisms in both models will differ.
In the second part of the report an efficient network with QoS capabilities is designed. To start with, the state-of-the-art or most used QoS mechanisms in the access, aggregation and core networks are identified. Next, it is argued why FTTC/VDSL2 and FTTH/P2P are taken as examples of NGN efficient networks. The last section of this part describes which main network elements must support QoS capabilities in FTTC/VDSL2 and FTTH/P2P networks in order to provide services with QoS.
Finally, the last part of the report discusses the following components of a cost model which could change when applying the cost model to a network that provides QoS: definition of cost drivers, cost of network elements, traffic carried out by network elements and routing factors. It is explained why the capacity of the network, measured in Kbps, Mbps or Gbps, could be the main cost driver. Another cost driver to take into account in nodes when using prioritization or capacity reservation is processing power. The deployment of network elements that support QoS will have an impact on the cost, and the traffic generated by QoS applications will change according to the QoS mechanisms employed. The routing path of traffic provided with QoS will also affect the usage factor of the network elements. Finally, the report includes examples of FTTC/VDSL2 and FTTH/P2P networks that illustrate the design of a routing matrix with usage factors for services provided with QoS.
Discussion Paper is available for download.