A new performance measure for statistical multiplexing: perspective of the individual user

Traditional methods of design of high-speed networks have focussed on performance measures such as probability of buffer overflow. Such a measure is likely to be inadequate for networks of the future because it does not reflect the performance seen by an individual user of the network. In this paper, we introduce the notion of perceptible degradation of quality of the received signal seen by an individual user. We propose that the design of high-speed networks that support real-time traffic be carried out on the basis of a new performance measure, which is defined as the expected number of perceptible degradations that a customer is likely to encounter per unit of time. By using fluid-flow models, we provide an analytical methodology for calculating the new performance measure. For data traffic, we also characterize performance in terms of the probability of burst loss. Based on this measure, we quantify the impact of two cell dropping policies, namely, the random policy and the selective policy, on the grade of service delivered to data applications. We illustrate the proposed methodology with representative numerical examples from voice, video and data traffic for various link speeds and buffer sizes.