We hear references to the importance of Quality of Service (QoS) in packet networks all the time. This concept is an important one as we increasingly use our packet networks to carry more than just data. Although QoS is not solely a matter of controlling delay, it is true that delay is an important part of the equation. Interactive voice is highly delay sensitive. Too much variation in delay (jitter) and the voice quality will be unacceptable. Too much delay, and the humans trying to talk struggle or the connection fails completely.

Before we can control delay, however, we have to understand it. Where does delay come from in a packet network? It turns out there are five primary sources of delay. In today's PodSnack, I'm going to walk you through them and explain why one of them is more complex to deal with than the others.

As always, you can listen to (or download) today's PodSnack by clicking here, or you can use the Flash player below.

If you have an idea or request for a PodSnack topic, send it to podsnacks@hill.com or simply put your request in a comment to this blog entry! You can get a complete list of our PodSnacks here. You can register for our PodSnack series by RSS feed by adding this link to your podcast aggregator: http://feeds.feedburner.com/HillAssociates-LivingInAConnectedWorldPodsnacks.