At a network-connected device, congestion at an egress queue can be detected. A potential source of congestion can be identified based on characteristics of a packet that caused the egress queue to become congested. The source of congestion can be a congestion group of transmitters. A group congestion message can be sent to the group of transmitters. The message can identify the packet that caused the egress queue to become congested. Transmitters can respond to the message by reducing their peak transmission rate.

A communication system in which DRA control is aided by RL. An example embodiment may control one or more buffer queues populated by downstream and/or upstream data streams. The egress rates of the buffer queues can be dynamically controlled using an RL technique, according to which a learning agent can adaptively change the state-to-action mapping function of the DRA controller while circumventing the RL exploration phase and relying on extrapolation of the already taken actions instead. This feature may result in at least two benefits: (i) cancellation of a performance penalty typically associated with RL exploration; and (ii) faster learning of the environment, as the learning agent can determine the performance metrics of many actions per state in a single occurrence of the state. In an example embodiment, the communication system may be a DSL system, a PON system, or a wireless communication system.

A communication system in which DRA control is aided by RL. An example embodiment may control one or more buffer queues populated by downstream and/or upstream data streams. The egress rates of the buffer queues can be dynamically controlled using an RL technique, according to which a learning agent can adaptively change the state-to-action mapping function of the DRA controller while circumventing the RL exploration phase and relying on extrapolation of the already taken actions instead. This feature may result in at least two benefits: (i) cancellation of a performance penalty typically associated with RL exploration; and (ii) faster learning of the environment, as the learning agent can determine the performance metrics of many actions per state in a single occurrence of the state. In an example embodiment, the communication system may be a DSL system, a PON system, or a wireless communication system.

Technologies for adaptive network packet egress scheduling include a switch configured to configure an eligibility table for a plurality of ports of the switch, wherein the eligibility table includes a plurality of rounds. The switch is further configured to retrieve an eligible mask corresponding to a round of a plurality of rounds of the eligibility table presently being scheduled and determine a ready mask that indicates a ready status of each port. The switch is further configured to determine, for each port, whether the eligible status and the ready status indicate that port is both eligible and ready, and schedule, in response to a determination that at least one port has been determined to be both eligible and ready, each of the at least one port that has been determined to be both eligible and ready. Additional embodiments are described herein.

A scheduling method applied in an industrial heterogeneous network in which a TSN and a non-TSN are interconnected is provided. The TSSDN controller classifies data flows according to the delay requirements, and calculates the scheduling priorities of the data flows in the industrial heterogeneous network. The TSSDN controller adopts an improved CSPF algorithm to determine a shortest path in the heterogeneous network, and marks the scheduling priorities of the data flows which are transmitted from the subnet of the heterogeneous network and arrive at the switch for the first time. Flow table matching is performed at the SDN switch. In a case of performing flow table matching successfully, the counter is updated and the instruction included in the flow table is executed. In a case of performing flow table matching unsuccessfully, a PacketIn message is transmitted to the TSSDN controller, and the TSSDN controller performs analysis and makes a decision.

Provided are a service flow transmission method and apparatus, a device, and a storage medium. The service flow transmission method includes: acquiring service flows; and performing transmission via different FlexE outgoing interfaces according to priorities of the service flows. By means of determining priorities of service flows, and performing transmission via different FlexE outgoing interfaces according to the priorities of the service flows, mutual interference between service flows can be prevented.

This application provides a buffer determining method and apparatus, to resolve a problem of how a terminal device on a sidelink calculates a buffer size. The method includes: A terminal device determines a sidelink data rate, and determines a buffer size based on the sidelink data rate. In this embodiment of this application, the terminal device may determine the buffer size based on the sidelink data rate, to calculate a buffer size of terminal device in sidelink communication.

This application provides a buffer determining method and apparatus, to resolve a problem of how a terminal device on a sidelink calculates a buffer size. The method includes: A terminal device determines a sidelink data rate, and determines a buffer size based on the sidelink data rate. In this embodiment of this application, the terminal device may determine the buffer size based on the sidelink data rate, to calculate a buffer size of terminal device in sidelink communication.

A network switch is capable of supporting cut-through switching and interface channelization with enhanced system performance. The network switch includes a plurality of ingress tiles, each tile including a virtual output queue (VOQ) scheduler operable to submit schedule requests to a fabric scheduler. Data is requested in unit of quantum, which may aggregate multiple packets, and which reduces schedule latency. Each request is associated with a start-of-quantum (SoR) state or a middle-of-quantum (MoR) state to support cut-through. The fabric scheduler performs a multi-stage scheduling process to progressively narrow the selection of requests, including stages of arbitration in virtual output port level, virtual output port group level, tile level, egress port level, and port group level. Each tile receives the grants for its requests and accordingly sends request data to a switch fabric for transmission to the destination egress ports.

Packets to be transmitted from a network device are buffered in queues in a first packet memory. In response to detecting congestion in a queue in the first packet memory, groups of multiple packets are transferred from the first packet memory to a second packet memory, the second packet memory configured to buffer a portion of traffic bandwidth supported by the network device. Prior to transmission of the packets among the one or more groups of multiple packets from the network device, packets among the one or more groups of multiple packets are transferred from the second packet memory back to the first packet memory. The packets transferred from the second packet memory back to the first packet memory are retrieved from the first packet memory and are forwarded to one or more network ports for transmission of the packets from the network device.