The L4S architecture enables internet applications to achieve low queuing latency, low congestion loss, and scalable throughput control. It proposes that the root cause of queuing delay is in the capacity-seeking congestion controllers of senders, rather than the queue itself. By adopting new congestion control algorithms that can seek capacity with minimal queuing, applications can have both low latency and high throughput. The architecture allows for incremental deployment, coexisting with existing congestion controls in a shared network. The three main components of the L4S architecture are the Scalable congestion control on the sending host, the Active Queue Management (AQM) at the network bottleneck, and the protocol between them.
https://datatracker.ietf.org/doc/rfc9330/