Voice quality network requirements
Available application models
Voice
The voice model converts low level statistics (network delay, latency, and loss) into an APR score that takes into account voice requirements. Given the sensitivity of voice to jitter, the voice model is especially sensitive to jitter. This model is also very sensitive to sustained loss, which effect can be detrimental to voice. This model is also sensitive to delay and loss. (See Section 3 for a detailed description of the effect of delay, jitter, and loss on voice quality).
Video conferencing
The video conferencing model converts low level statistics into an Application Performance Rating (APR) score that takes into account video conferencing requirements. Given the detrimental effects of loss on video conferencing, this model is very sensitive to loss. This model is also sensitive to jitter and latency.
Enterprise model
The enterprise model takes into account the effect of low level network characteristics on short TCP transactions. This model is very sensitive to network loss. On the other hand, this model is not affected by jitter. Given that signaling traffic uses TCP and consists of short transactions, The CNA application model that best captures the characteristics of signaling traffic is the enterprise application model.
Web model
The web model is best suited for web applications. It converts latency and loss measures into an APR score that best describes the user experience in the context of a typical web transaction.
For more information on CNA, see The Converged Network Analyzer on page 344.
