What is a description of the DiffServ model used for implementing QoS?

In Assured Forwarding (AF), a higher drop precedence number (e.g., AF43) indicates a higher likelihood of being dropped than a lower drop precedence number (e.g., AF41), meaning AF41 has lower drop precedence than AF43.

In the DiffServ model, different types of traffic are classified and marked with specific DiffServ Code Point (DSCP) values. Voice and video calls, having different latency and jitter requirements, are typically assigned distinct DSCP values (e.g., EF for voice, AF41-AF43 for video) and then placed into separate queues based on these markings for differentiated handling.

AF43 has higher drop precedence (more likely to be dropped) than AF42, and AF42 has higher drop precedence than AF41; therefore, AF43 has higher drop precedence than AF41.

While RTP traffic from voice is marked EF, RTP traffic from video is typically marked with an Assured Forwarding (AF) class (e.g., AF41) or a different DSCP value, not EF, because voice and video have different characteristics and tolerances.