Training a software agent to control a communication network

    公开(公告)号:US11177996B2

    公开(公告)日:2021-11-16

    申请号:US16498680

    申请日:2017-04-04

    摘要: A method for training a Software Agent to control a communication network is disclosed. The method comprises initialising a training network slice within the communication network, instantiating within the training network slice a cloned version of at least one Virtualised Network Function (VNF) comprised within a production network slice, mirroring traffic incoming to the VNFs of the production network slice and forwarding the mirrored traffic to the training network slice. The method further comprises causing a training instance of the Software Agent to execute a Reinforcement Learning algorithm on the training network slice, and transferring knowledge acquired by the training instance of the Software Agent to a production instance of the Software Agent.
    Also disclosed are an apparatus and a computer program configured to carry out methods for training a Software Agent to control a communication network.

    Management of analytics tasks in a programmable network

    公开(公告)号:US10887178B2

    公开(公告)日:2021-01-05

    申请号:US16083543

    申请日:2016-03-17

    摘要: Distributed management of analytics tasks in a programmable network (100) having a controller (SDNC, 10, 11), a plurality of network nodes (20, N1-N5) coupled to the controller by a Data Communications Network DCN, and a plurality of analytics servers (30) each coupled to a network node, involves the controller selecting which of the analytics servers to use for an analytics task. The selection is based on how data logged by respective ones of the network nodes and needed for that analytics task, can be routed to the analytics servers. The controller sends to the selected analytics server, an indication of the analytics task and an indication of what data logged by the network nodes is to be used in the analytics task.

    Group recommendation for selection of service items

    公开(公告)号:US10204370B2

    公开(公告)日:2019-02-12

    申请号:US15032117

    申请日:2013-10-28

    摘要: The present disclosure provides enhanced group recommendation for selection of a service item with respect to a service class, and communicates the enhanced group recommendation to a sub-group of service users. Initial service scores for service items in the service class are generated by aggregating individual service preferences of sub-group members with respect to the service class. Group preferences for service items in the service class are then predicted using statistic data on previous service selections of sub-group members with respect to the service class. The initial service scores are aggregated with the predicted group preferences to generate of weighted service scores with respect to service items in the service class. One of the service items in the service class that has the highest weighted score is the group recommendation is communicated to the sub-group members.

    Method and User Equipment for Predicting Available Throughput for Uplink Data

    公开(公告)号:US20170202000A1

    公开(公告)日:2017-07-13

    申请号:US15315279

    申请日:2014-06-04

    IPC分类号: H04W72/08 H04L5/00 H04W72/04

    摘要: A method in a user equipment for predicting an available throughput for uplink data to be sent from the user equipment to a base station over a radio link. The user equipment obtains (201) an information about a past scheduling of the user equipment. The user equipment obtains (202) a radio quality measure related to the radio link. The user equipment predicts (204) the available throughput based on a relationship between the information about a past scheduling of the user equipment and one or more previously obtained pieces of information about the past scheduling of the user equipment. The user equipment predicts the available throughput further based on a relationship between the radio quality measure and of one or more previously obtained radio quality measures. The user equipment predicts the available throughput further based on a previously obtained throughput. The previously obtained throughput is associated with the one or more previously obtained pieces of information about a past scheduling and the one or more previously obtained radio quality measures.

    Adaptive cell selection in heterogeneous networks

    公开(公告)号:US10264496B2

    公开(公告)日:2019-04-16

    申请号:US15326605

    申请日:2014-08-15

    摘要: A method of performing cell selection for a set of mobile terminals among at least two neighboring cells including a first and a second radio access node, respectively, includes acquiring a value of at least one property reflecting a network condition for each one of the mobile terminals, determining with which one of the first and second radio access nodes each of the mobile terminals initially should be associated based on the value of the acquired property for each mobile terminal, and generating a plurality of variations to the determined initial association based on a corresponding selected parameter. The method includes performing cell selection for at least a subset of the generated plurality of variations to the initial association, and determining a value of a metric for each of the performed cell selections and associating each metric value with said acquired property.

    Efficient operation of user equipment in a wireless communication network

    公开(公告)号:US10172182B2

    公开(公告)日:2019-01-01

    申请号:US15030239

    申请日:2013-11-07

    摘要: A method of operating a User Equipment, UE, in a wireless communication network. The UE is configured for Discontinuous Reception, DRX, operation with a DRX cycle having a DRX ON period in which the UE is in a DRX awake state and ready for reception and a DRX OFF period in which the UE is in a DRX sleep state. The UE determines whether an uplink packet to be transmitted is delay tolerant. The UE delays, when the uplink packet is regarded as delay tolerant, a request for an uplink transmission grant for the uplink packet until a DRX ON period of the UE. By delaying the request for an uplink transmission grant for a delay tolerant uplink packet until a DRX ON period of the UE, the UE does not have to wake up to listen to the response from the network during the DRX OFF period.

    Network node and a method therein for computing cell range expansion (CRE) values

    公开(公告)号:US10154417B2

    公开(公告)日:2018-12-11

    申请号:US15500450

    申请日:2014-09-08

    摘要: A method in a network node for computing a Cell Range Expansion (CRE) value. The network node obtains a first value of a measure of the performance of the wireless communications network. The network node creates a first relationship relating the first value of the parameter to the first CRE value and the first measure of the performance of the wireless communications network, based on the first value of the parameter and the first value of the measure of the performance. The network node creates a second relationship relating the second value of the parameter to a second CRE value and a second value of the measure of the performance of the wireless communications network, based on the first relationship and on the second value of the parameter. The network node computes the second CRE value based on a second value of the parameter, and the second relationship.

    Method and network node for pre-scheduling grant transmission

    公开(公告)号:US10064213B2

    公开(公告)日:2018-08-28

    申请号:US15104636

    申请日:2013-12-16

    摘要: A pre-scheduling grant transmission method performed by a network node includes issuing a first sequence of pre-scheduling grants to provide a UE with reserved resources for uplink data packet transmission from a send buffer of the UE, collecting data relating to at least transmitted uplink traffic initiated by the first pre-scheduling grants to obtain a data set providing a statistical description of the traffic flow during the time period covered by the first sequence of pre-scheduling grants, predicting from the obtained data set the time when the next data packet arrives to the send buffer of the UE, and transmitting at least one pre-scheduling grant to the UE at a time in the vicinity of the predicted time, thereby providing the UE with timely resources for transmitting the data packet shortly after the packet has arrived at the UE send buffer.

    Methods, wireless device and network node for managing positioning method based on prediction

    公开(公告)号:US09749805B2

    公开(公告)日:2017-08-29

    申请号:US14917642

    申请日:2013-10-25

    摘要: A method and a wireless device (110) for managing a positioning method based on prediction as well as a method and a network node (120) for managing mobility patterns for use by a wireless device (110) when performing a positioning method based on prediction are disclosed. The wireless device (110) performs mobility measurements relating to mobility of the wireless device (110). The wireless device (110) estimates a position of the wireless device (110) based on prediction by using the mobility measurements and a first intra-cell pattern. The wireless device (110) detects whether the wireless device (110) is on the first intra-cell pattern or on a second intra-cell pattern, when the estimated position has reached a checkpoint of the first intra-cell pattern. The network node (120) receives information about mobility patterns. The mobility patterns comprise intra-cell patterns of cells visited by the wireless device (110). The network node (120) identifies one or more checkpoints by finding a junction between at least two of the intra-cell patterns.

    Network node, a base station for uplink grant pre-allocation

    公开(公告)号:US10219289B2

    公开(公告)日:2019-02-26

    申请号:US15026487

    申请日:2013-10-01

    摘要: A method in a network node downloads an upcoming web page to a user equipment. The network node obtains a first value, relating to the upcoming web page, of a set of features. The features represent any one out of: the upcoming web page, the user equipment, and a web browser in the user equipment. The network node predicts whether it is beneficial regarding download time to activate an uplink grant pre-allocation for the user equipment for downloading the upcoming web page. The prediction is based on a relationship between the first value and of a second value of the set of features, which second value is related to a previously downloaded web page, and an evaluation whether it is beneficial regarding download time to activate the uplink grant pre-allocation for the previously downloaded web page. The network node provides the prediction to a base station.