A constant revenue model for packet switched network

摘要

It is well known that the quality of service as measured by latency in a packet switched network degrades rapidly when the incident traffic exceeds the design limits of the network. A higher latency would result in customer dissatisfaction and likely large scale customer flight to alternative service providers with the attendant loss of revenue to the service provider. This paper presents a mechanism that introduces a priority system with the objective of providing a higher and a lower quality of service to the two customer groups. The non-priority traffic carries a lower price tag and a lower quality of service. An important characteristic of the proposed pricing schemes is that the overall revenue associated with the network remains constant as long as the total demand is confined within a relatively large bound, termed the region of operation, for the network.