Embodiments of this application disclose a network slice management method and device, and relate to the field of communications technologies. The method includes: receiving, by a first management unit, a network slice management request, where the network slice management request carries instance information or indication information of a transport network manager; and sending, by the first management unit, a transmission management request to a corresponding transport network manager based on the instance information of the transport network manager, where the transmission management request is used to deploy a transmission network. The embodiments of this application provide a method for determining a transport network manager to create a corresponding transmission network.

Certain aspects and features of the present disclosure relate to systems and methods that generate machine-learning models to predict whether user devices are likely to meet defined objectives. For example, a machine-learning model can be generated to predict whether or not a user device is likely to access a resource. In some implementations, a semi-supervised model can be used to determine to what extent user devices are predicted to satisfy the defined objective(s). For example, a resource-affinity parameter can be generated as a result of inputting various data points into a semi-supervised model. The various data points can be access from a plurality of data sources, and can represent one or more activities or attributes associated with a user. The value of the resource-affinity parameter can be evaluated to determine the extent to which the user is likely to meet an objective.

Certain aspects and features of the present disclosure relate to systems and methods that generate machine-learning models to predict whether user devices are likely to meet defined objectives. For example, a machine-learning model can be generated to predict whether or not a user device is likely to access a resource. In some implementations, a semi-supervised model can be used to determine to what extent user devices are predicted to satisfy the defined objective(s). For example, a resource-affinity parameter can be generated as a result of inputting various data points into a semi-supervised model. The various data points can be access from a plurality of data sources, and can represent one or more activities or attributes associated with a user. The value of the resource-affinity parameter can be evaluated to determine the extent to which the user is likely to meet an objective.

Technologies for dynamic accelerator selection include a compute sled. The compute sled includes a network interface controller to communicate with a remote accelerator of an accelerator sled over a network, where the network interface controller includes a local accelerator and a compute engine. The compute engine is to obtain network telemetry data indicative of a level of bandwidth saturation of the network. The compute engine is also to determine whether to accelerate a function managed by the compute sled. The compute engine is further to determine, in response to a determination to accelerate the function, whether to offload the function to the remote accelerator of the accelerator sled based on the telemetry data. Also the compute engine is to assign, in response a determination not to offload the function to the remote accelerator, the function to the local accelerator of the network interface controller.

A controller device includes a memory configured to store a tree structure comprising a plurality of nodes, wherein the tree structure comprises a set of sub-structures, and wherein the tree structure defines a configuration of a network device of a set of network devices such that each node of the plurality of nodes corresponds to a respective resource of the network device. Additionally, the controller device includes processing circuitry configured to receive an instruction to update the configuration of the network device, wherein the instruction to update the configuration of the network device indicates a node of the set of nodes corresponding to the update; and verify, based on a sub-structure of the set of sub-structures corresponding to the node indicated by the instruction, the instruction to update the configuration of the network device.

Technologies for managing configuration-free platform firmware include a compute device, which further includes a management controller. The management controller is to receive a system configuration request to access a system configuration parameter of the compute device and access the system configuration parameter in response to a receipt of the system configuration request.

According to an example aspect of the present invention, there is provided a method comprising, receiving, from a first data endpoint of a radio access network, a representation of a local model of the first data endpoint of the radio access network, determining multiple common models for endpoints of the radio access network, selecting, based on the representation of the local model of the first data endpoint, one of said multiple common models for the first data endpoint and transmitting the selected common model to the first data endpoint, any other data endpoint or any other external system which utilizes the selected common model.

A computing node to implement an RL management entity in an NG wireless network includes a NIC and processing circuitry coupled to the NIC. The processing circuitry is configured to generate a plurality of network measurements for a corresponding plurality of network functions. The functions are configured as a plurality of ML models forming a multi-level hierarchy. Control signaling from an ML model of the plurality is decoded, the ML model being at a predetermined level (e.g., a lowest level) in the hierarchy. The control signaling is responsive to a corresponding network measurement and at least second control signaling from a second ML model at a level that is higher than the predetermined level. A plurality of reward functions is generated for training the ML models, based on the control signaling from the MLO model at the predetermined level in the multi-level hierarchy.

A method including receiving, from a device management element or function of at least one transport and/or access device or function installed at a predefined location for which a local control is to be conducted, and storing device or function related data, forwarding the stored device or function related data to a centralized control element or function, receiving, from the centralized control element or function, and processing instruction data for the at least one transport and/or access device or function, and conducting a local device or function management control procedure for the at least one transport and/or access device or function according to a result of the processing of the instruction data.

A software defined network (SDN) system, controller, and controlling method, where the SDN system includes at least one N.sup.th level controller and at least two (N+1).sup.th level controllers belonging to the N.sup.th level controller, where the (N+1).sup.th level controller is configured to receive a first message sent by a node belonging to the (N+1).sup.th level controller, and when the first message is a cross-domain message according to status information of each node that is managed by the (N+1).sup.th level controller, forward the first message to the N.sup.th level controller to which the (N+1).sup.th level controller belongs, and the N.sup.th level controller receives the first message, and perform decision processing according to status information of the (N+1).sup.th level controller that is managed by and belongs to the N.sup.th level controller and status information of boundary nodes of the (N+1).sup.th level controller belonging to the N.sup.th level controller.