Embodiments of this application disclose a radio frequency resource allocation method, apparatus, device, system, and a storage medium. The method includes: obtaining radio frequency information of an access point AP (for example, RSSI signal strength between the AP and each neighboring AP, and data traffic of the AP in a data collection period), predicting, based on the radio frequency information of the AP, load of the AP that is in target duration after a current moment, and allocating a radio frequency resource to the AP based on the load of the AP. This implementation can reduce actual interference on an entire network and improve user experience.

A network device includes first, second, and third processors. The first processor detects congestion in a packet flow. The packet flow is i) one packet flow among a plurality of packet flows and ii) is formed of a plurality of packets of a same type received from a first device in a network via a first network connection. The packets in the packet flow are destined for a second device in the network. When congestion notification packet generation is enabled for the packet flow, the second processor generates a congestion notification packet by replicating a packet from the packet flow and sends the congestion notification packet to the first device via the first network connection. The congestion notification packet identifies the packet flow for which congestion is detected. The third processor forwards the plurality of packets in the packet flow to the second device via a second network connection.

A network device includes first, second, and third processors. The first processor detects congestion in a packet flow. The packet flow is i) one packet flow among a plurality of packet flows and ii) is formed of a plurality of packets of a same type received from a first device in a network via a first network connection. The packets in the packet flow are destined for a second device in the network. When congestion notification packet generation is enabled for the packet flow, the second processor generates a congestion notification packet by replicating a packet from the packet flow and sends the congestion notification packet to the first device via the first network connection. The congestion notification packet identifies the packet flow for which congestion is detected. The third processor forwards the plurality of packets in the packet flow to the second device via a second network connection.

Graded throttling for network-on-chip traffic, including: calculating, by an agent of a network-on-chip, a number of outstanding transactions issued by the agent; determining that the number of outstanding transactions meets a threshold; and implementing, by the agent, in response to the number of outstanding transactions meeting the threshold, a traffic throttling policy.

Graded throttling for network-on-chip traffic, including: calculating, by an agent of a network-on-chip, a number of outstanding transactions issued by the agent; determining that the number of outstanding transactions meets a threshold; and implementing, by the agent, in response to the number of outstanding transactions meeting the threshold, a traffic throttling policy.

This application discloses a network congestion control method and a network device, and relates to the field of communication technologies, to improve continuity and completion of a packet in a transmission process of a Transmission Control Protocol (TCP) flow. In this application, when pre-congestion (that is, congestion may occur) occurs in a forwarding direction of the TCP flow, a forwarding device reduces a value of a receive window (RWND) field in an acknowledge (ACK) packet, to reduce a sending rate of a transmit end for a packet in the TCP flow or delay sending of a packet by a transmit end. This relieves buffer pressure of the forwarding device, reduces a packet loss, and improves the continuity and the completion of the packet in the transmission process of the TCP flow.

This application discloses a network congestion control method and a network device, and relates to the field of communication technologies, to improve continuity and completion of a packet in a transmission process of a Transmission Control Protocol (TCP) flow. In this application, when pre-congestion (that is, congestion may occur) occurs in a forwarding direction of the TCP flow, a forwarding device reduces a value of a receive window (RWND) field in an acknowledge (ACK) packet, to reduce a sending rate of a transmit end for a packet in the TCP flow or delay sending of a packet by a transmit end. This relieves buffer pressure of the forwarding device, reduces a packet loss, and improves the continuity and the completion of the packet in the transmission process of the TCP flow.

The present disclosure describes solutions for dynamic network slicing including provisions to create, modify, and/or delete network slices in a de-centralized communication network including a plurality of central/regional/edge/far-edge locations across hybrid and multi-cloud environment referred to as edge server or edge location for providing service to the users. Network slicing enables multiple isolated and independent virtual (logical) networks to exist together. A plurality of virtual networks, i.e., slices, may be created using resources of the same physical network infrastructure.

The present disclosure describes solutions for dynamic network slicing including provisions to create, modify, and/or delete network slices in a de-centralized communication network including a plurality of central/regional/edge/far-edge locations across hybrid and multi-cloud environment referred to as edge server or edge location for providing service to the users. Network slicing enables multiple isolated and independent virtual (logical) networks to exist together. A plurality of virtual networks, i.e., slices, may be created using resources of the same physical network infrastructure.

In one embodiment, a device predicts a failure of a first tunnel in a software-defined wide area network (SD-WAN). The device determines that no backup tunnel for the first tunnel exists in the SD-WAN that can satisfy one or more service level agreements (SLAs) of traffic on the first tunnel, were the traffic rerouted from the first tunnel onto that tunnel. The device predicts, using a machine learning model, that a backup tunnel for the first tunnel exists in the SD-WAN that can satisfy an SLA of a subset of the traffic on the first tunnel, in response to determining that no backup tunnel exists in the SD-WAN that can satisfy the one or more SLAs of the traffic on the first tunnel. The device proactively reroutes the subset of the traffic on the first tunnel onto the backup tunnel, in advance of the predicted failure of the first tunnel.