Speech traffic is often used in simulations to evaluate the performance of control mechanisms in communication networks. Therefore, trustworthy models are required that capture the fundamental statistical properties of typical voice sources. The G.723.1 codec produces on/off traffic streams with fixed size packets. The iSAC codec strongly periodic packet streams with variable packet sizes. We propose new models for the traffic output of both codecs and show that their queuing properties are in good accordance with those of original traffic traces, while existing traffic models, that are frequently used in literature, lead to significant discrepancies.
展开▼
