/mcpQuality of Service for Voice over IP Fragmentation and Interleaving 10 QoSVoIP.mif Fragmentation and Interleaving Because VoIP transmissions are extremely delay-sensitive, VoIP packets must be interleaved or inserted between data packet fragments. This section describes fragmentation and interleaving, and includes the following subsections: • Fragmentation and Interleaving Overview • MLP Link Fragmentation and Interleaving Example • FRF.12 Fragmentation and Interleaving Example Fragmentation and Interleaving Overview Even if queueing is working at its best and prioritizing voice traffic, there are times when the priority queue is empty and a packet from another class is serviced. Packets from guaranteed bandwidth classes must be serviced according to their configured weight. If a priority voice packet arrives in the output queue while these packets are being serviced, the VoIP packet could wait a substantial amount of time Class-Based WFQ (CBWFQ) MQC is used to classify traffic. Classified traffic is placed into reserved bandwidth queues or a default unreserved queue. A scheduler services the queues based on weights so that the bandwidth guarantees are honored. Similar to LLQ except that there is no priority queue. Simple configuration and ability to provide bandwidth guarantees. No priority servicing is possible. Priority Queue WFQ (PQ-WFQ, also called IP RTP Priority) A single interface command is used to provide priority servicing to all UDP packets destined to even port numbers within a specified range. Simple, one command configuration. Provides priority servicing to RTP packets. All other traffic is treated with WFQ. RTCP traffic is not prioritized. No guaranteed bandwidth capability. LLQ (Previously called PQ-CBWFQ) MQC is used to classify traffic. Classified traffic is placed into a priority queue, reserved bandwidth queues, or a default unreserved queue. A scheduler services the queues based on weights so that the priority traffic is sent first (up to a certain policed limit during congestion) and the bandwidth guarantees are met. Simple configuration. Ability to provide priority to multiple classes of traffic and give upper bounds on priority bandwidth utilization. You can also configure bandwidth guaranteed classes and a default class. No mechanism for providing multiple levels of priority yet—all priority traffic is sent through the same priority queue. Separate priority classes can have separate upper priority bandwidth bounds during congestion, but sharing of priority queue between applications may introduce jitter. Table 1 Software Queueing Mechanisms (continued) Software Queueing Mechanism Description Benefits Limitations