3. Context

NQB traffic is low-rate and smooth enough that it should not materially create queues, even at inter-packet timescales. QB traffic, by contrast, includes bursty or high-rate microflows driven by classic congestion control and tends to create queuing delay and loss.

NQB fits the Diffserv architecture as a PHB that distinguishes observable traffic behavior rather than application importance. It is not a premium priority service. The incentive design depends on NQB being a better queue for compliant NQB traffic and the Default/QB queue being the better place for traffic that builds queues.

The relationship to L4S is complementary. L4S traffic can use scalable congestion control and ECN, while NQB identifies non-queue-building behavior. NQB network functions SHOULD treat NQB packets marked ECT(1) or with other ECN field values consistently, as later required in Section 5.4.

The most important deployment target is access-network bottlenecks with relatively deep buffers, including cable broadband, Wi-Fi, and mobile access segments. Lower-rate links need extra care because the sender-rate assumptions may not hold.