Abstract: There is provided a method performed by a central entity of a network. A first set of features is selected for a machine learning model to take into account when analysing data. The machine learning model is to be deployed at an edge entity of the network. The selection is based on first information indicative of data that is available for the machine learning model to analyse, second information indicative of features that are available for the machine learning model to take into account when analysing data, and contextual information associated with the network.

Abstract: A method is provided for determining bias of machine learning models. The method includes: forming a training dataset including input data samples provided to a remote machine learning model developed using a machine learning process, and corresponding output data samples obtained from the remote machine learning model; training a local machine learning model which approximates the remote machine learning model using a machine learning process and the training dataset; and interrogating the trained local machine learning model to determine whether the remote machine learning model is biased with respect to one or more biasing data parameters.

Abstract: A method performed by a node in a communications network for determining a target network configuration for use in providing services to a first operator on the communications network. The method includes obtaining characteristics of the first operator; obtaining, for a plurality of previous operators in the communications network, characteristics of the previous operators and corresponding target network configurations used for the respective previous operators; matching the first operator to a second operator selected from the previous operators based on similarity between the characteristics of the first operator and the characteristics of the previous operators; and determining the target network configuration for the first operator, based on a target network configuration used for the second operator.

Abstract: A computer implemented method performed by a node in a communications network comprises obtaining Key Performance Indicator, KPI, targets for a plurality of KPIs, in the communications network, and determining relationships between the KPIs using a model trained using a graph-based machine learning process. Each relationship describes a manner in which changing a network configuration to alter a first one of the plurality of KPIs affects a second one of the plurality of KPIs. The method then comprises determining one or more conflicts between the KPI targets, using the relationships.

Abstract: A computer implemented method performed by a node in a communications network for matching parameters in a target network configuration with corresponding key performance indicators, KPIs, in the communications network. The method comprises obtaining (202) a first set of connections between a first subset of the parameters and a first subset of the KPIs. The method further comprises representing (204) the parameters as a first knowledge graph, representing (206) the KPIs as a second knowledge graph, and using (208) a graph-based machine learning process on the first knowledge graph and the second knowledge graph to match a second parameter to a second KPI.

Abstract: A method, in a first network function, for providing one or more platoon recommendations for platooning of a plurality of communication devices the method comprising: obtaining mobility information for the plurality of communication devices; determining, based on the mobility information, a platoon recommendation comprising a subset of the plurality of communication devices to be grouped into a platoon and a head communication device of the subset of communication devices to be responsible for reporting reference signal measurements on behalf of the platoon; and transmitting the platoon recommendation to a second network function.

Abstract: Mechanisms for migration of a stateless computation service in a mobile communication network. A method, performed by a control node comprises obtaining, from a core network node in the mobile communication network, information indicative of a need to migrate the stateless computation service from a source compute node in the mobile communication network. The need results from resource utilization at the source compute node is too high, and/or a demand from user equipment for the stateless computation service elsewhere in the mobile communication network. The method comprises initiating migration of the stateless computation service from the source compute node to a target compute node in the mobile communication network for hosting of the stateless computation service at the target compute node.

Abstract: Embodiments described herein relate to methods and apparatuses for training a neural network. A method comprises receiving an input data set at a layer of the neural network; performing a forward pass and a backward pass on the input data set to determine regular output data; calculating a first loss associated with the regular output data; performing a quantized forward pass and a quantized backward pass on the input data set to determine quantized output data; calculating a second loss associated with the quantized output data; comparing the first loss to the second loss; and based on the comparison determining whether to reduce the input data set to provide a reduced data set.

Abstract: The present disclosure relates to a method of scheduling communication of a wireless communication device in a communications network, and a device performing the method. A method of a device of scheduling communication of a wireless communication device in a communications network is provided. The method includes determining a geographical path along which the wireless communication device is expected to travel, acquiring data indicating power availability of access points via which the wireless communication device is capable of communicating along the expected geographical path, determining, from the acquired data indicating power availability, via which of the access points communication is to be established along the expected geographical path for avoiding establishing communication with one or more of the access points indicated to be subjected to power failure, and scheduling the communication to occur via the determined access points.

Abstract: A computer implemented method performed by a node in a communications network comprises obtaining Key Performance Indicator, KPI, targets for a plurality of KPIs, in the communications network, and determining relationships between the KPIs using a model trained using a graph-based machine learning process. Each relationship describes a manner in which changing a network configuration to alter a first one of the plurality of KPIs affects a second one of the plurality of KPIs. The method then comprises determining one or more conflicts between the KPI targets, using the relationships.

Abstract: Methods and central nodes for developing machine-learning models, the method including receiving, at a central node, at least one episode including a plurality of changes of an environment. The method further includes analysing the episode to extract observations and grouping the observations from among the plurality of observations into a plurality of groups of similar observations. A first machine learning agent is then trained using a first group of similar observations from among the plurality of groups of similar observations, and a second machine learning agent is trained using a second group of similar observations from among the plurality of groups of similar observations, wherein the second group of similar observations is different to the first group of similar observations. The central node obtains a central machine-learning model based on an output from at least one of the trained first machine learning agent and the trained second machine learning agent.

Abstract: A testing system and method for testing a Printed Circuit Boards (PCBs) is provided. The method is being executed at a testing system which comprises a RF test analyzer; a RF energy source; and one or more RF probes. The method includes performing a first level scanning of a first set of components in the PCB. The method further includes performing a second level scanning of another set of components in the PCB, which differs from the first set of components in the PCB; the second level scanning is performed only if anomalies are identified which is based on analyzing the results of the performed first level scanning. The method further includes determining detailed root causes of the identified anomalies which is based on analyzing results of the performed second level scanning.

Abstract: A method performed by a content server (100) for determining additional media content to be inserted into a main media content wherein the main media content is in a virtual environment, the method comprising: acquiring (S201) information relating to a user profile of each of a plurality of users that are viewing the main media content; clustering (S202) the plurality of users into one or more user clusters based on the user profile of each of the plurality of users; acquiring (S203) a viewing state of each of the plurality of users that are viewing the main media content, wherein the viewing state of each of the plurality of users comprises biometric data of each of the plurality of users; acquiring (S204) information relating to a network quality; and determining (205) the additional media content for each of the plurality of users by a first reinforcement learning agent based on the one or more user clusters, the viewing state of each of the plurality of users, and the network quality.

Abstract: AM and LM parameters to be used for adapting an ASR model are derived for each audio segment of an audio stream comprising multiple audio programs. A set of identifiers, including a speaker identifier, a speaker domain identifier and a program domain identifier, is obtained for each audio segment. The set of identifiers are used to select most suitable AM and LM parameters for the particular audio segment. The embodiments enable provision of maximum constraints on the AMs and LMs and enable adaptation of the ASR model on the fly for audio streams of multiple audio programs, such as broadcast audio. This means that the embodiments enable selecting AM and LM parameters that are most suitable in terms of ASR performance for each audio segment.

Abstract: Methods and devices enable connecting devices to cellular networks using the devices' hardware identifiers. Subscriber records include a hardware identifier assigned when the devices are manufactured. A target hardware identifier included in an attachment request is associated with an International Mobile Subscriber Identity, IMSI, available to the cellular network if, according to subscriber records, the device is registered.

Abstract: A reinforcement learning (RL) process is used to allocate UL/DL resources and is also used to offload traffic from a first access point (e.g., macro access point) to a second access point (e.g., a micro access point) when the load on the first access point is too high. The RL process is able to handle the dynamic nature of UL/ DL imbalances and is therefore able to maximize usage of resources without compromising quality of service.

Abstract: Reporting information included in data records generated for network traffic activities of a plurality of mobile devices is described. A data record includes a traffic type of a traffic activity of a mobile device, a timestamp of the traffic activity, and an identifier of the mobile device. A method is provided that includes receiving a monitoring request specifying a monitoring type, and calculating from the data records a number of mobile devices having, within a given monitoring period, traffic activities that fulfil the following conditions: i) a traffic type of the traffic activity matches the monitoring type; ii) a timestamp of the traffic activity falls within the monitoring period; and iii) traffic activities associated with the same identifier and the same traffic activity are considered only once per monitoring period. The method further comprises returning, a monitoring report that is based on the calculated number of mobile devices.

Abstract: A method (200) performed by a node (100) in a telecommunications network for selecting a preferred action to take from a plurality of proposed actions in order to achieve a target network configuration in the telecommunications network. The method comprises obtaining (202) the plurality of proposed actions, evaluating (204) each of the plurality of proposed actions compared to the target network configuration using a computational argumentation process, and selecting (206) the preferred action, based on the results of the evaluating.

Abstract: Methods, a wireless device (110), a radio network node (120) and a core network node (130) for reservation of bandwidth are disclosed. The wireless device (110) sends (A010), to the core network node (130), an inquiry for reservation of bandwidth. The inquiry comprises information about a route to be travelled by the wireless device (110), a start time, and an indication about the bandwidth to be reserved. The core network node (130) sends (A040), to the radio network node (120), a command for reservation of the bandwidth to be reserved for the wireless device (110). The command comprises the indication, information relating to one or more cells of the radio network node (120), in which one or more cells bandwidth is to be reserved according to the indication, a parameter indicating a period of time during which the wireless device (110) is expected to make use of at least some of the bandwidth to be reserved, and an identity of the wireless device (110).

Abstract: AM and LM parameters to be used for adapting an ASR model are derived for each audio segment of an audio stream comprising multiple audio programs. A set of identifiers, including a speaker identifier, a speaker domain identifier and a program domain identifier, is obtained for each audio segment. The set of identifiers are used to select most suitable AM and LM parameters for the particular audio segment. The embodiments enable provision of maximum constraints on the AMs and LMs and enable adaptation of the ASR model on the fly for audio streams of multiple audio programs, such as broadcast audio. This means that the embodiments enable selecting AM and LM parameters that are most suitable in terms of ASR performance for each audio segment.