LONG-RANGE DEPENDENCE IN VBR VIDEO STREAMS AND ATM TRAFFIC ENGINEERING

Studies at Bellcore have shown that traffic rates in high-speed data n etworks exhibit 'long-range dependence' in their correlation, characte rized by the so-called Hurst parameter H taking values in the range 0. 5 < H < 1. This corresponds to a slower asymptotic decay of correlatio n than in finite-state Markovian processes, for which H = 0.5. While i t has been shown that, in some situations, link-engineering based on M arkovian traffic models could underestimate the bandwidth requirements for traffic with H > 0.5, there are other results showing the success ful use of a Markovian traffic model to determine bandwidth requiremen ts for variable-bit-rate video sources, which do have H > 0.5. We repo rt here the results of an investigation undertaken to bridge the gap i n our understanding of these two sets of results, and to examine the e ffect of the Hurst parameter on link-engineering for ATM traffic. Our investigation has shown that a curious 'scaling' property of streams w ith long-range dependence, derived in another recent Bellcore study, h elps to bridge the gap by accounting for both sets of results. The ana lysis suggests that a high value of H is not, in itself, a reason to s uppose that Markovian traffic models will lead to under-engineering of bandwidth on ATM links. (C) 1997 Elsevier Science B.V.