Radio parameter allocation

Abstract

A method to allocate radio parameters to a user equipment (UE) in a wireless network using a trained model. Radio parameter data indicative of a first setting being used by the UE is received, 202. Quality of experience (QoE) data indicative of a quality of a user experience is also received from the UE, 204. The radio parameter data and the QoE data are provided to a parameter allocation model, 206, which determines a second setting of a radio parameter for the UE such that a prescribed QoE is met. The new radio parameter setting is then sent to the UE, 208. The radio parameter allocation model receives a plurality of training data records. Each record comprises QoE, measurement data indicative of a quality of a user experience provided by the UE whilst using a historical setting of a radio parameter. The radio parameter may be related to the configuration of: Hybrid Automatic Repeat Request (HARQ), discontinuous reception, discontinuous transmission, Medium Access Control protocol, Radio Link Control (RLC) protocol or Packet Data Convergence Protocol (PDCP). The QoE data may comprise metrics such as throughput, playout delay, buffer level, perceptual quality, jitter or relate to a user's experience streaming media content.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to methods and apparatus for allocating radio parameters to be used by a user equipment (UE) attached to a wireless communication network. The method and apparatus may find particular application in allocating radio parameters for use by the UE and sending the allocated radio parameters to the UE using a Radio Resource Control (RRC) message. BACKGROUND

[0002] A wireless communication network, such as a mobile telecommunications network, provides a range of services to devices attached to the wireless communication network. For example, an electronic device attached to the network may be provided with voice, messaging and / or data services. Such services are provided by exchanging radio frequency signals (radio signals) between the electronic device and a network component such as a base station. In order to transmit and / or receive radio signals over the wireless communication network a device may be allocated one or more radio parameters to be used by the device. For example, a device may be allocated one or more radio configuration parameters to be used when transmitting and / or receiving radio signals over the network.

[0003] Radio parameters are typically managed by a network component (such as a base station) so as to utilise the available radio frequency spectrum and network resources as efficiently as possible whilst minimising interference. The network component may allocate radio parameters to a plurality of different devices attached to the network and communicate the allocated parameters to the devices for use in future exchange of radio signals over the network. For example, allocated radio configuration parameters may be communicated to a device by sending messages using a Radio Resource Control (RRC) protocol.

[0004] Users are increasingly using devices attached to wireless networks for high bit-rate real time services over the network, such as video streaming services or extended reality (XR) services. For these types of services, even intermittent quality degradation can have a large impact on the quality of a user's experience with the service. Many of these streaming services are a significant part of growth in network traffic in wireless communication networks. It may be desirable to improve the end users' experiences when accessing services using a wireless communication network.

[0005] It is in this context the present disclosure has been devised BRIEF SUMMARY

[0006] It has been realised that the allocation of radio parameters in a wireless communication network can influence the quality of a user experience when receiving a service using the wireless communication network. For example, an optimal allocation of radio parameters to different devices attached to the network may serve to improve a quality of user experience provided at one or more devices. Methods and apparatus are presented herein for allocating radio configuration parameters in a wireless communication network which accounts for a quality of experience whilst providing a service using the network. In particular, data which is indicative of a quality of user experience may be collected and used to allocate radio configuration parameters in such a way as to improve a quality of user experience.

[0007] According to a first aspect of the present disclosure there is provided a method of allocating radio parameters for a first user equipment (UE) in a wireless communication network. The method comprises receiving radio parameter data indicative of a first setting of at least one radio parameter being used for the first UE whilst the first UE provides a service using the wireless communication network. The method further comprises receiving quality of experience (QoE) data from the first UE, wherein the QoE data is indicative of a quality of a user experience provided by the first UE whilst the first UE provides the service using the wireless communication network and using the first setting for the at least one radio parameter. The method further comprises providing the received radio parameter data and the received QoE data to a parameter allocation model, wherein the parameter allocation model is configured to determine, in dependence on the radio parameter data and the QoE data, a second setting of the at least one radio parameter for use for the first UE in the wireless communication network and for the first UE to receive at least a prescribed QoE whilst using the second setting of the at least one radio parameter. The method further comprises sending the second setting of the at least one radio parameter, determined by the parameter allocation model, to the first UE.

[0008] It has been found that by allocating a setting of the at least one radio parameter based, at least on part on the QoE data, then the QoE at the first UE can be improved in a way which was not possible using previous methods of allocating settings of radio parameters. For example, by relating settings of radio parameters directly to a QoE at a UE, a setting for one or more radio parameters may be changed specifically in order to an improve a QoE and / or to address a problem in a QoE which is identified by QoE data. [00091 The QoE data may relate to a quality of a user experience which may be linked to a specific end user service or end user service type (e.g., playout and / or streaming of media content). QoE data may, for example, relate to factors which are directly observable by a user whilst receiving an end user service (e.g., delay, quality, etc. experienced when receiving a service such as playout of media content). QoE data is measured at a device which provides the service. That is, the QoE data is measured at a UE. The QoE data may be measured by a UE which decodes data and provides a service (e.g., displays a service). QoE data is typically measured and / or collected at an application layer (e.g., by an application providing an end user service). QoE data differs from other measures of network performance such as, for example, radio signal measurements and / or Quality of Service (QoS) parameters. Radio signal measures are not typically measured and directly experienced at an application layer but are determined in a radio layer. QoS parameters are determined at a service layer based on a service parameter. These QoS parameters may be propagated to the radio layer to be considered in radio parameter setting at the radio layer. Some QoS parameter may also be measured at a radio layer.

[0010] The QoE data may be measured and / or collected by the first UE. For example, the QoE data may be collected by an end user application operating in the UE. The QoE data may be provided from an application layer to a radio layer in which the methods described herein may be implemented. In some examples, QoE data may be inferred indirectly from measures made in a radio layer. For example, one or more QoE metrics may be mapped onto radio layer measurements (for example, using a relationship derived from historical observations) so as to determine QoE data at least in part indirectly from radio layer measurements.

[0011] The QoE data may comprise one or more metrics such as an average throughput, an initial playout delay, a buffer level, a playlist, a playout delay for media start-up, a perceptual quality, jitter, an intermediate playout delay and device information. The QoE data may relate to a quality of a user's experience whilst streaming of media content using the wireless communication network.

[0012] QoE data may be indicative of a quality of experience over a period of time. For example, QoE metrics may be collected over a profile period of time over which a QoE profile is evaluated. The profile period of time may be greater than 1 millisecond and may, for example, be a period of a few tens of milliseconds. QoE data may therefore be indicative of a quality of experience which has occurred over a profile period of time. [00131 The first setting of the at least one radio parameter may comprise a setting of the at least one radio parameter which was used through a profile period of time to which the QoE data relates. In some examples, a setting of the at least one radio parameter may have been changed during the profile period of time to which the QoE data relates. In such examples, the first setting may comprise each setting of the at least one radio parameter which was used through the profile period of time. That is, the first setting may comprise a plurality of settings.

[0014] The at least one radio parameter may comprise at least one radio parameter which is used to configure an aspect of the way in which data is received by and / or transmitted by the first UE using the wireless communication network. For example, the at least one radio parameter may comprise at least one radio parameter which configures a procedure to be used by the first UE and / or a Radio Access Network (RAN) node when exchanging data (uplink data and / or downlink data). The second setting of the at least one radio parameter which is sent to the first UE may allow the first UE to configure a procedure for exchanging data using the wireless communication network according to the first setting.

[0015] The parameter allocation model may be configured to determine the second setting of the at least one radio parameter for use for the first UE for the first UE to receive at least a prescribed QoE whilst using the second setting of the at least one radio parameter and for at least another UE to receive at least a prescribed QoE whilst the second setting is used for the first UE. That is, the second setting may be determined for the first UE in dependence on its impact on a QoE provided at the first UE and on its impact on a QoE provided at the least another UE. The at least another UE may comprise a plurality of other UEs. The at least another UE may be attached to the wireless communication network.

[0016] As was described above, the parameter allocation model is configured to determine, the second setting of the at least one radio parameter for the first UE to receive at least a prescribed QoE. The prescribed QoE may represent a minimum QoE which is desired or required for the first UE. The prescribed QoE may be referred to as a required QoE. The prescribed QoE may comprise a pre-defined measure of QoE. The prescribed QoE may be dynamic. For example, the prescribed QoE may be updated dependent on an application, service or task being performed or provided by the first UE. The prescribed QoE may comprise a value of at least one metric representative of a QoE. For example, the prescribed QoE may comprise a value for at least one metric including one or more of an average throughput, an initial playout delay, a buffer level, a playlist, a playout delay for media startup, a perceptual quality, jitter and / or intermediate playout delay. It will be appreciated that an improvement in a QoE may correspond with an increase in some metrics of QoE (such as an average throughput) but may correspond with a decrease in other metrics of QoE (such as a buffer level and / or playout delay). The prescribed QoE may comprise a minimum of a first metric for which an increase represents an improvement (such as an average throughput) and / or may comprise a maximum of a second metric for which a decrease represents an improvement (such as a buffer level and / or playout delay).

[0017] The parameter allocation model may be configured to determine the second setting of the at least one radio parameter for use for the first UE for the first UE to receive an improved QoE whilst using the second setting. That is, the parameter allocation model may be configured to determine the setting second of the at least one radio parameter so as to bring about an improvement in the QoE provided at the first UE. The improvement in the QoE may be an improvement relative to a QoE currently being provided whilst using a current setting of the at least one radio parameter. For example, the improvement in QoE may be an improvement relative to a QoE indicated by the received QoE data. As was explained above, an improvement in a QoE may comprise an increase in a value of one or more first metrics (such as an average throughput) and / or may comprise a decrease in a value of one or more second metrics (such as a buffer level and / or playout delay).

[0018] The parameter allocation model may be configured to determine the second setting of the at least one radio parameter for use for the first UE to bring about an overall improvement in a QoE provided by a plurality of UEs attached to the wireless communication network. For example, the second setting may be determined in order to bring about an improvement in a QoE provided at a plurality of UEs which includes the first UE.

[0019] The parameter allocation model may be configured through training. For example, the parameter allocation model may be configured through training based on training data records comprising data related to settings of the at least one radio parameter which have been used in the network and QoE data which indicates a QoE which resulted from the use of the setting of the at least one radio parameter.

[0020] The parameter allocation model may comprise an artificial intelligence and / or machine learning model. The parameter allocation model may comprise a model configured through training using an artificial intelligence and / or machine learning method. The parameter allocation model may be configured through training using a plurality of training data records. The plurality of training data records may comprise historical QoE data indicative of a QoE provided at a plurality of UEs in the wireless communication network. The plurality of training data records may further comprise historical settings of the at least one radio parameter which were used when the QoE indicated by the historical QoE data was received.

[0021] The parameter allocation model may be configured through training using a supervised learning process. The parameter allocation model may be configured through training using a reinforcement learning process.

[0022] The parameter allocation model may be configured for a given geographical region (e.g. a region comprising one or more cells), for a given period of time (such as a given time of day and / or time of year), for given weather conditions (e.g. for fine, cloudy, rainy weather conditions etc.), and / or for given network load conditions. For example, different parameter allocation models may be configured to one or more different regions, time periods and / or conditions. In at least some examples, different parameter allocation models may be configured through training based on different training data records, where the different training data records correspond to different regions, different time periods and / or different conditions. Different parameter allocation models may be used to determine the second setting of the at least one radio parameter in different regions, for different time periods and / or in different conditions.

[0023] Sending the second setting of the at least one radio parameter, determined by the parameter allocation model, to the first UE may comprise sending a control message (e.g., a control plane message) to the first UE. The control message may include the second setting of the at least one radio parameter and may be configured to cause the first UE to use the second setting of the at least one radio parameter for data transmission and / or reception when using the wireless communication network. The control message may comprise a Radio Resource Control (RRC) message.

[0024] The method of allocating radio parameters for a first user equipment may be performed when it is determined that allocating a new setting of the at least one radio parameter may provide an improvement in a QoE at the first UE.

[0025] The method of allocating radio parameters may be a computer implemented method. The method may be performed by any suitable one or more computing devices. In some examples, the method may be implemented at least in part at a Radio Access Network (RAN) node such as a base station. The RAN node may comprise a RAN node which operates a serving cell for the first UE.

[0026] The wireless communication network may comprise a cellular network and / or a mobile telecommunications network. [00271 The at least one radio parameter may comprise a plurality of radio parameters. The first setting and the second setting may comprise settings of each of the plurality of radio parameters.

[0028] The at least one radio parameter may comprise a set of radio parameters for use by the first UE. The first and second settings of the at least one radio parameter may comprise a setting for each of the set of radio parameters. For example, the first setting of the at least one radio parameter may comprise a setting for a first radio parameter and a setting for a second radio parameter. The first setting of the at least one radio parameter may further comprise a setting for a third radio parameter and may further comprise further settings for one or more further radio parameters. Similarly, the second setting of the at least one radio parameter may comprise a setting for a first radio parameter and a setting for a second radio parameter. The second setting of the at least one radio parameter may further comprise a setting for a third radio parameter and may further comprise settings for one or more further radio parameters.

[0029] The at least one radio parameter may comprise at least one radio parameter for which a setting of the radio parameter is allocated for use for the first UE for a plurality of instances of data transmission and / or reception.

[0030] An allocated setting of the at least one radio parameter may not be allocated for each instance of data transmission and / or reception. A setting of the at least one radio parameter may be allocated for use (and e.g. used for) transmitting and / or receiving a plurality of transport blocks.

[0031] The at least one radio parameter may comprise at least one radio parameter for which a setting of the radio parameter is allocated for use for the first UE for a plurality of transmission time intervals.

[0032] The at least one radio parameter may comprise at least one radio parameter for which a setting of the radio parameter is typically used for the first UE for a time period which is greater than about 500 milliseconds. The at least one radio parameter may comprise at least one radio parameter for which a setting of the radio parameter is typically used for the first UE for a time period which is greater than about 1 second.

[0033] Sending the second setting of the at least one radio paramater to the first UE may comprise sending the second setting to the first UE as part of a Radio Resource Control, RRC, message.

[0034] An RRC message may comprise a message using an RRC protocol and / or a message which is sent as part of an RRC procedure. Sending the second setting of the at least on radio parameter to the first UE may comprise sending the second setting to the first UE as part of an RRC procedure. Examples of RRC procedures (during which an RRC message may be sent including the second setting of the at least one radio parameter) may include (but are not necessarily limited to) a connection establishment procedure, a connection release procedure, a connection maintenance procedure, a cell handover procedure, a context transfer procedure, a cell selection and / or re-selection procedure, an inter-Radio Access Technology (RAT) mobility procedure, a paging procedure, and / or an outer loop power control procedure.

[0035] The at least one radio parameter may comprise at least one of a parameter related to the configuration of a Hybrid Automatic Repeat Request procedure, a parameter related to the configuration of a discontinuous reception procedure, a parameter related to the configuration of a discontinuous transmission procedure, a parameter related to the configuration of a procedure carried out using a Medium Access Control protocol, a parameter related to the configuration of a procedure carried out using a Radio Link Control protocol, and / or a parameter related to the configuration of a procedure carried out using a Packet Data Convergence Protocol.

[0036] Examples of parameters related to the configuration of a Hybrid Automatic Repeat Request (HARQ) procedure may include (but are not necessarily limited to) a maximum number of HARQ processes to be carried out for each data transmission.

[0037] Examples of parameters related to the configuration of a discontinuous reception (DRX) procedure may include (but are not necessarily limited to) a duration of a DRX cycle (e.g., DRX cycle), a duration of “ON” time within one DRX cycle (e.g. OnDurationTimer), a duration of time that a UE should remain “ON” after reception of a Physical Downlink Control Channel (PDCCH) message (e.g., drx-Inactivity timer), a maximum number of consecutive PDCCH subframes that a UE should remain active to wait for an incoming re-transmission (e.g., drx-Retransmission timer), a parameter related to a short DRX cycle which can be implemented (e.g., shortDRX cycle) and / or a duration of time that a UE should use a short DRX cycle (e.g., drxShortCycleTimer).

[0038] Examples of parameters related to the configuration of a discontinuous transmission (DTX) procedure may include one or more parameters which configure the timing and / or procedures to be used in a DTX cycle.

[0039] Examples of parameters related to the configuration of a procedure carried out using the Medium Access Control (MAC) protocol may include (but are not necessarily limited to) one or more parameters used to configure MAC parameters for a cell group. For example, parameters related to the configuration of a procedure carried out using the MAC protocol may include one or more of a parameter which sets the use of a pre-emptive Buffer Status Report (BSR), a parameter which limits the sending of a Channel State Information (CSI) report to an on-duration period of a DRX cycle, a timer after which a connection is released if there is data inactivity for duration of the timer, a parameter for configuring a DRX procedure, a parameter which sets a prioritisation of a MAC entity for overlapping grants and scheduling requests, a parameter which indicates a scheduling request configuration applicable for a beam failure recovery on a secondary cell, a parameter which indicates a scheduling request configuration applicable for consistent uplink Listen-Before-Talk (LBT) recovery on a secondary cell, a parameter which determines skipping of uplink transmissions, and / or a parameter used to configure a time-alignment group. [00401 Examples of parameters related to the configuration of a procedure carried out using the Radio Link Control (RLC) protocol may include (but are not necessarily limited to) one or more parameters for setting an RLC mode and / or for configuring a radio bearer such as a signalling radio bearer and / or a data radio bearer. For example, parameters related to the configuration of a procedure carried out using the RLC protocol may include one or more of a parameter indicated whether to operate using an RLC transparent mode (TM), acknowledged mode (AM) or unacknowledged mode (UM), a parameter which sets a maximum number of transmissions for use in RLC acknowledged mode, a number of bytes for which polling of peers should be performed in RLC acknowledged mode, a number of packet data units (PDU) for which polling of peers should be performed in RLC acknowledged mode, a sequence number field size, a timer for a transmitting side for retransmitting a poll in RLC acknowledged mode, a timer for a receiving side to detect loss of RLC PDUs at lower layers in RLC acknowledged mode, and / or a timer for a receiving side to prohibit transmission of a STATUS PDU in RLC acknowledged mode.

[0041] Examples of parameters related to the configuration of a procedure carried out using the Packet Data Convergence Protocol (PDCP) may include (but are not necessarily limited to) one or more configurable parameters for signalling and data radio bearers. For example, parameters related to the configuration of a procedure carried out using the PDCP protocol may include one or more of a parameter for disabling ciphering, a timer for discarding a service data unit (SDU) and a corresponding PDU at a transmitting PDCP entity, a parameter indicated whether a PDCP entity continues or resets a robust header compression (ROHC) protocol during PDCP re-establishment, a parameter indicating a PDCP uplink duplication state for an associated RLC entity, a parameter configuring ethernet header compression, a parameter configuring ROHC, a parameter indicating whether or not integrity protection is configured for a radio bearer, a parameter indicating a maximum value for a Context Identifier (CID), a parameter which configures uplink data transmission when more than one RLC entity is associated with the PDCP entity, a parameter which sets whether a PDCP SDU is delivered to upper layers after performing header decompression, a parameter which indicates whether or not uplink duplication status is configured and activated, a PDCP sequence number size for downlink, a PDCP sequence number size for uplink, a parameter which indicates a cell group identification and a logical channel identification (LCID) of a primary RLC entity, a parameter indicating an LCID of a split secondary RLC entity, a parameter indicating whether a data radio bearer is configured to send a PDCP status report in the uplink, and / or a timer used to detect loss of PDCP data PDUs, a threshold amount of data for transmitting data to a primary and secondary RLC entity if a split secondary RLC entity is configured. [00421 The parameter allocation model may be configured through training based on a plurality of training data records. Each training data record may comprise: a historical setting of the at least one radio parameter and measured QoE data. The measured QoE data may be indicative of a quality of a user experience provided by a UE whilst the UE provided a service using the wireless communication network and using the historical setting for the at least one radio parameter.

[0043] The historical setting of the at least one radio parameter may comprise a setting of the at least one radio parameter which has been allocated to a UE connected to the wireless communication network. The at least one radio parameter may comprise a plurality of radio parameters. For example, the historical setting of the at least one radio parameter may comprise a setting for a first radio parameter and a setting for a second radio parameter. The historical setting of the at least one radio parameter may further comprise a setting for a third radio parameter and may further comprise a setting for one or more further radio parameters.

[0044] The measured QoE data is associated with an historical setting of the at least one parameter which was in place at the time of the measurement of QoE. Put another way, the historical setting of the at least one radio parameter represents the setting of the at least one radio parameter which was being used at the time at which the QoE indicated by the measured QoE data was experienced.

[0045] The measured QoE data may comprise a value of at least one metric representative of a QoE. For example, the measured QoE may comprise a value for at least one metric including one or more of an average throughput, an initial playout delay, a play list, a buffer level, a playout delay for media start-up, a perceptual quality, jitter and / or intermediate playout delay. [00461 The training data records may comprise data which was collected by and / or measured by a plurality of different UEs operating in the wireless communication network. For example, the training data records may comprise data which was collected by and / or measured by one or more UEs different to the first UE.

[0047] The training data records may further comprise other data and / or measurements. For example, the training data records may further comprise measured radio coverage data indicative of radio coverage provided to a UE whilst the UE provides the service using the wireless communication network and using the historical setting for the at least one radio parameter. The measured radio coverage may for example comprise one or more of a measure of a received signal power (e.g. a received reference signal power (RSRP)), a received signal quality (e.g. a received reference signal quality (RSRQ)), a signal to noise ratio (e.g. a signal to interference and noise ratio (SINR)), and / or a channel quality indicator (e.g. channel state information (CSI)). The radio coverage data may be associated with a serving cell and / or one or more neighbouring cells. For example, the measured radio coverage data may comprise one or more of a measure of a received signal power (e.g. RSRP), a received signal quality (e.g. RSRQ), a signal to noise ratio (e.g. SINR), and / or a channel quality indicator (e.g. CSI) associated with a serving cell. Additionally or alternatively the measured radio coverage data may comprise one or more of a measure of a received signal power (e.g. RSRP), a received signal quality (e.g. RSRQ), a signal to noise ratio (e.g. SINR), and / or a channel quality indicator (e.g. CSI) associated with a serving cell associated with one or more neighbouring cells.

[0048] The parameter allocation model may comprise a predictive model, configured through training to determine a predicted QoE which is expected to result from an input setting of the at least one radio parameter.

[0049] The predictive model may be trained using a supervised learning process. The supervised learning process may be based on training data records, which may have any of the properties of training data records which were described above.

[0050] A predicted QoE may comprise a value of at least one metric representative of a QoE. For example, the predicted QoE may comprise a value for at least one metric including one or more of an average throughput, an initial playout delay, a buffer level, a playlist, a playout delay for media start-up, a perceptual quality, jitter and / or intermediate playout delay.

[0051] The predictive model may be configured to determine a predicted QoE associated with a plurality of UEs in the wireless communication network. For example, the predictive model may be configured to determine a predicted QoE for the first UE and for one or more other UEs attached to the wireless communication network. In such examples, the method may further comprise receiving a setting of the at least one radio parameter and QoE data for the one or more other UEs. The method may further comprise providing a setting of the at least one radio parameter and QoE data for the one or more other UEs to the predictive model.

[0052] The parameter allocation model may be further configured to determine the second setting of the at least one radio parameter as a setting of the at least one radio parameter which the predictive model expects to result in at least a prescribed QoE whilst providing a service at the first UE and using the second setting of the at least one radio parameter.

[0053] The parameter allocation model may for example, be configured to generate one or more potential input settings of the at least one radio parameter. The predictive model may determine a predicted QoE which is expected to result from each of the one or more potential input settings of the at least one radio parameter. The parameter allocation model may be configured to select from the one or more potential input settings to determine the second setting of the at least one radio parameter. The selection may be based on the predicted QoE which is expected to result from each of the one or more potential input settings. For example, the parameter allocation model may select as the second setting, a setting from the one or more potential input settings which is associated with an expected QoE which corresponds to at least the prescribed QoE. The parameter allocation model may select as the second setting, a setting from the one or more potential input settings which is associated with an expected QoE which represents an improvement in QoE. The parameter allocation model may select as the second setting, a setting from the one or more potential input settings which is associated with an expected QoE which represents a greatest improvement in QoE (e.g., from a plurality of expected improvements in QoE associated with a plurality of potential input settings).

[0054] The parameter allocation model may select as the second setting, a setting from the one or more potential input settings which is associated with an expected QoE for a plurality of UEs which corresponds to at least the prescribed QoE. The parameter allocation model may select as the second setting, a setting from the one or more potential input settings which is associated with an expected QoE which represents an improvement in QoE for a plurality of UEs. The parameter allocation model may select as the second setting, a setting from the one or more potential input settings which is associated with an expected QoE which represents a greatest overall improvement in QoE for a plurality of UEs (e.g., from a plurality of expected improvements in QoE associated with a plurality of potential input settings). [00551 The parameter allocation model may be configured through training comprising a reinforcement learning process. Determining the second setting of the at least one radio parameter may comprise an iteration of the reinforcement learning process.

[0056] The reinforcement learning process may comprise a deep reinforcement learning process. The reinforcement learning process may comprise a multi-agent reinforcement learning (MARL) process. In such examples, different UEs attached to the wireless communication network may have different agents associated with them. That is, each agent in a MARL process may be associated with a different UE. A MARL process may be carried out for a plurality of UEs so as to allocate radio parameters to each of the plurality of UEs as part of the same MARL process.

[0057] A reinforcement learning process generally takes a series of actions and observes changes in states which result from the actions. Through such a process, a reward which is associated with different actions is learned in order to choose future actions which maximise or optimize the reward. When applied to examples disclosed herein the actions in a reinforcement learning process may comprise an allocation of a setting for the at least one radio parameter. A state which is observed in a reinforcement learning process may comprise the received QoE data which is associated with the first setting (a given action) of the at least one radio parameter. A reward in a reinforcement learning process may comprise an improvement in at least one metric of a QoE for one or multiple UEs. When applied to examples disclosed herein, in each iteration of a reinforcement learning process an action in the form of a second setting of the at least one radio parameter may be determined in order to seek a reward in the form of an improvement in at least one metric of a QoE for one or multiple UEs.

[0058] In some examples, a centralized training with decentralized execution process may be carried out. In such examples, the parameter allocation model may be trained using data collected for a plurality of UEs in the network but may be executed independently for different UEs. In some examples, centralized or co-operative learning may be used. For example, a plurality of Radio Access Network (RAN) nodes, such as base stations, may provide data corresponding to UEs attached to each RAN node in order to perform a centralized learning process. In some examples a co-operative learning process may comprise co-operation between different RAN nodes.

[0059] The reinforcement learning process may be configured to maximise a reward comprising at least a measure of a QoE provided by the first UE whilst the first UE provides a service using the wireless communication network. [00601 In some examples, the reward may comprise at least one metric of a QoE provided by the first UE. For example, the reward may comprise a measure for at least one metric including one or more of an average throughput, an initial playout delay, a buffer level, a playout delay for media start-up, jitter and / or intermediate playout delay. It will be appreciated that an improvement in a QoE may correspond with an increase in some metrics of QoE (such as an average throughput) but may correspond with a decrease in other metrics of QoE (such as a buffer level and / or playout delay). Maximising (or at least increasing) a reward comprise at least one measure of a QoE may, for example, comprise maximizing (or at least increasing) at least a first metric for which an increase represents an improvement (such as an average throughput) and / or may comprise a minimizing (or at least decreasing) at least a second metric for which a decrease represents an improvement (such as a buffer level and / or playout delay).

[0061] In some examples, the reward may comprise at least one metric of a QoE provided by a plurality of UEs which include the first UE. For example, the reinforcement learning process may be configured to improve or optimize QoE for a plurality of UEs.

[0062] In some examples, the reward may be a function of a QoE and one or more other factors. For example, the reward may further comprise a measure of one or more other properties such as efficient use of resources (which may include radio resources and / or computing resources), radio coverage and / or load balancing in the network.

[0063] The method may further comprise: receiving second radio parameter data indicative of a plurality of settings of the at least one radio parameter being used for a plurality of second UEs whilst the plurality of second UEs provide a service using the wireless communication network. The method may further comprise receiving second quality of experience, QoE, data from the second UEs, wherein the second QoE data is indicative of a quality of a user experience provided by the second UEs whilst the second UEs provide the service using the wireless communication network and using the plurality of settings for the at least one radio parameter. The method may further comprise providing the second received radio parameter data and the second received QoE data to the parameter allocation model, wherein the parameter allocation model is configured to determine the second setting of the at least one radio parameter for use by the first UE in further dependence on the second received radio parameter data and the second received QoE data.

[0064] In at least some examples, the parameter allocation model may be configured to determine the second setting so as to provide a prescribed QoE both for the first UE and for one or more second UEs also connected to the wireless communication network. In such examples, the parameter allocation model may be provided with inputs which relate both to the first UE and the one or more second UEs. The inputs may, for example, comprise settings of the at least one radio parameter and corresponding QoE data which corresponds to a time period in which the settings of the at least one radio parameter were used. In this way, the parameter allocation model is provided with data illustrative of the current QoE and the current radio parameter settings which are in use in the network and may use that data to determine the second setting of the at least one radio parameter for use for the first UE.

[0065] In some examples, the parameter allocation model may be configured to allocate settings of the at least radio parameters for use by a plurality of UEs. For example, the parameter allocation model may be configured to determine optimum or improved settings of the at least one radio parameter for a group of UEs. In some examples, the parameter allocation model may be configured to determine a plurality of second settings of the at least one radio parameter, where the plurality of second settings comprise a setting for each of a plurality of UEs. For example, the parameter allocation model may determine a second setting of the at least one radio parameter for use by the first UE and second settings of the at least one radio parameter for use by each of the one or more second UEs.

[0066] The parameter allocation model may be configured to determine the second setting of the at least one radio parameter for use for the first UE and the second plurality of UEs to receive at least a prescribed QoE whilst the second setting of the at least one radio parameter is used for the first UE.

[0067] The plurality of second UEs may include at least some UEs attached to a plurality of Radio Access Network nodes of the wireless communication network.

[0068] In some examples, the parameter allocation model determines the second setting of the at least one radio parameter based on data corresponding to UEs which are attached to a plurality of different cells. For example, data may be received which corresponds to one or more UEs which are attached to a serving cell for the first UE and / or may data may be received which corresponds to one or more UEs which are attached to one or more neighbouring cells. In this way the parameter allocation model is provided with data which is indicative of a QoE across a geographical area which spans more than one cell.

[0069] The method may further comprise receiving radio coverage data from the first UE, wherein the radio coverage data is indicative of radio coverage provided to the first UE whilst the first UE provides the service using the wireless communication network and using the first setting for the at least one radio parameter. The parameter allocation model may be configured to determine the second setting of the at least one radio parameter for use by the first UE in further dependence on the received radio coverage data.

[0070] The received radio coverage data may comprise and / or may be based on one or more measurements made by the first UE. The radio coverage data may, for example, comprise one or more a measure of a received signal power (e.g. a received reference signal power (RSRP)), a received signal quality (e.g. a received reference signal quality (RSRQ)), a signal to noise ratio (e.g. a signal to interference and noise ratio (SINR)), and / or a channel quality indicator (e.g. channel state information (CSI)). The radio coverage data may be associated with a serving cell and / or one or more neighbouring cells. For example, the radio coverage data may comprise one or more of a measure of a received signal power (e.g. RSRP), a received signal quality (e.g. RSRQ), a signal to noise ratio (e.g. SINR), and / or a channel quality indicator (e.g. CSI) associated with a serving cell. Additionally or alternatively the radio coverage data may comprise one or more of a measure of a received signal power (e.g. RSRP), a received signal quality (e.g. RSRQ), a signal to noise ratio (e.g. SINR), and / or a channel quality indicator (e.g. CSI) associated with one or more neighbouring cells.

[0071] The method may further comprise performing a cell handover for the first UE, wherein the cell handover comprises handover of the first UE from a first cell operated by a first Radio Access Network (RAN) node to a second cell operated by a second RAN node. The method may further comprise sending the determined second setting of the at least one radio parameter to the second RAN node for use by the second RAN node.

[0072] The method may be performed by the first RAN node, which may operate a serving cell for the first UE. Under some conditions the first UE may undergo a handover to a cell operated by the second RAN node. In order to assist the second RAN node in allocating a suitable setting of the at least one radio parameter, the second RAN node may be provided with the setting (the second setting) which has been allocated and / or used for the first UE by the first RAN node. The parameter allocation model and / or one or more parameters of the parameter allocation model may be further send to the second RAN node.

[0073] According to a second aspect of the present disclosure there is provided a method of training a radio parameter allocation model for determining a setting of at least one radio parameter for use for a UE in a wireless communication network. The method comprises receiving a plurality of training data records. Each training data record comprises: an historical setting of the at least one radio parameter used for a UE in the wireless communication network; and Quality of Experience (QoE) measurement data indicative of a quality of a user experience provided by the UE whilst the UE provided a service using the wireless communication network and using the historical setting of the at least one radio parameter. The method further comprises training the parameter allocation model, in dependence on the received training data records, for determining a setting of at least one radio parameter for use for a UE in the wireless communication network and for the UE to receive at least a prescribed QoE whilst using the determined setting.

[0074] Any of the features and / or properties described above with reference to a method of allocating radio parameters may also apply to a method of training a radio parameter allocation model. For example, the QoE measurement data may have any of the properties described above with reference to QoE data received as part of a method of allocating radio parameters.

[0075] According to a third aspect of the present disclosure there is provided a computer-readable medium comprising instructions which, when executed by a computing apparatus, cause the computing apparatus to carry out any method described herein (e.g., a method according to the first aspect and / or a method according to the second aspect).

[0076] According to a fourth aspect of the present disclosure there is provided a computing system for use in a wireless communication network. The computing system is configured to receive radio parameter data indicative of a first setting of at least one radio parameter being used for a first UE whilst the first UE provides a service using the wireless communication network. The computing system is further configured to receive quality of experience (QoE) data from the first UE, wherein the QoE data is indicative of a quality of a user experience provided by the first UE whilst the first UE provides the service using the wireless communication network and using the first setting for the at least one radio parameter. The computing system is further configured to provide the received radio parameter data and the received QoE data to a parameter allocation model, wherein the parameter allocation model is configured to determine, in dependence on the radio parameter data and the QoE data, a second setting of the at least one radio parameter for use for the first UE in the wireless communication network and for the first UE to receive at least a prescribed QoE whilst using the second setting of the at least one radio parameter. The computing system is further configured to send the second setting of the at least one radio parameter, determined by the parameter allocation model, to the first UE.

[0077] According to a fifth aspect of the present disclosure there is provided a computing system for use in a wireless communication network. The computing system is configured to: receive a plurality of training data records. Each training data record comprises: an historical setting of the at least one radio parameter used for a UE in the wireless communication network; and Quality of Experience (QoE) measurement data indicative of a quality of a user experience provided by the UE whilst the UE provided a service using the wireless communication network and using the historical setting of the at least one radio parameter. The computing system is further configured to train a parameter allocation model, in dependence on the received training data records, for determining a setting of at least one radio parameter for use for a UE in the wireless communication network and for the UE to receive at least a prescribed QoE whilst using the determined setting.

[0078] It will be appreciated from the foregoing disclosure and the following detailed description of the examples that certain features and implementations described as being optional in relation to any given aspect of the disclosure set out above should be understood by the reader as being disclosed also in combination with the other aspects of the present disclosure, where applicable. Similarly, it will be appreciated that any attendant advantages described in relation to any given aspect of the disclosure set out above should be understood by the reader as being disclosed as advantages of the other aspects of the present disclosure, where applicable. That is, the description of optional features and advantages in relation to a specific aspect of the disclosure above is not limiting, and it should be understood that the disclosures of these optional features and advantages are intended to relate to all aspects of the disclosure in combination, where such combination is applicable. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] Certain examples of the present disclosure will now be described, with reference to the accompanying drawings, in which: - FIG. 1 is a schematic illustration of an example of a wireless communication network; - FIG. 2 is a flowchart of an example method of allocating radio parameters for a user equipment; - FIG. 3 is a flowchart of example method of the operation of a radio parameter allocation model which comprises a predictive model; - FIG. 4 is a flowchart of an example method of training a radio parameter allocation model for determining a setting of at least one radio parameter for use for a user equipment; - FIG. 5 is a signal flow diagram of an example method of allocating radio parameters for a user equipment; - FIG. 6 is a signal flow diagram of further steps of an example method of allocating radio parameters for a user equipment; and - FIG. 7 is a schematic illustration of an example electronic device which may be used to implement all or some of the methods described herein DETAILED DESCRIPTION

[0080] Hereinafter, examples of the disclosure are described with reference to the accompanying drawings. However, it should be appreciated that the disclosure is not limited to the described examples, and all changes and / or equivalents or replacements thereto also belong to the scope of the disclosure. The same or similar reference denotations may be used to refer to the same or similar elements throughout the specification and the drawings.

[0081] As used herein, the terms “have,” “may have,” “include,” or “may include” a feature (e.g., a number, function, operation, or a component such as a part) indicate the existence of the feature and do not exclude the existence of other features. Throughout the description and claims of this specification, the words “comprise” and “contain” and variations of them mean “including but not limited to”, and they are not intended to (and do not) exclude other components, integers or steps. Throughout the description and claims of this specification, the singular encompasses the plural unless the context otherwise requires. In particular, where the indefinite article is used, the specification is to be understood as contemplating plurality as well as singularity, unless the context requires otherwise.

[0082] As used herein, the terms “A or B,” “at least one of A and / or B.” or “one or more of A and / or B” may include all possible combinations of A and B. For example, “A or B,” “at least one of A and B,” “at least one of A or B” may indicate all of (1) including at least one A, (2) including at least one B, or (3) including at least one A and at least one B.

[0083] As used herein, the terms “first” and “second” may modify various components regardless of importance and do not limit the components. These terms are only used to distinguish one component from another. For example, reference to a first component and a second component may indicate different components from each other regardless of the order or importance of the components.

[0084] It will be understood that when an element (e.g., a first element) is referred to as being (physically, operatively or communicatively) “coupled with / to,” or “connected with / to” another element (e.g., a second element), it can be coupled or connected with / to the other element directly or via a third element. In contrast, it will be understood that when an element (e.g., a first element) is referred to as being “directly coupled with / to” or “directly connected with / to” another element (e.g., a second element), no other element (e.g., a third element) intervenes between the element and the other element.

[0085] The terms as used herein are provided merely to describe some embodiments thereof, but not to limit the scope of other embodiments of the disclosure. It is to be understood that the singular forms “a,” “an,” and “the” include plural references unless the context clearly dictates otherwise. All terms including technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the embodiments of the disclosure belong. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the relevant art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.

[0086] As used herein, "base station" refers to a network node comprised in a wireless communication network (which can also be referred to as a Radio Access Network (RAN)) for providing, via one or more antennae, cells of radio coverage in which a wireless radio connection to user equipment may be established. The Base Station may comprise a base station (BS), radio base station, base transceiver station (BTS), base station controller (BSC), radio network controller (RNC), evolved Node B (cNB or eNodeB), Node B, next generation eNodeB (ng-eNodeB), gNodeB (gNB), multi-standard radio (MSR) radio node such as MSR BS, multi-cell / multicast coordination entity (MCE), relay node, donor node controlling relay, radio access point (AP), and suchlike. Each base station provides one or more cells of a radio communications system within which wireless radio connections can be established between the base stations and user equipment. Each base station may support voice and data communication by one or more Radio Access Technologies (RATs), such as UMTS, LTE, 5G NR and 6G. The cells provided by each base station are roughly designated as land areas of coverage arranged as sectors emanating from radio antennas arranged at the base station. User equipment located in the cells in which coverage is provided can gain network access by establishing a radio connection with the base station providing the cell. The Radio Access Network can provide access through a Core Network to other networks such as the public Internet, and can allow voice and data communication with other user equipment in the Radio Access Network.

[0087] As used herein, "wireless communication network" refers to a network of base stations arranged to support radio communications within one or more cells emanating therefrom implementing one or more Radio Access Technologies to establish wireless radio connections to user equipment to enable them to communicate with the core network and other connected networks via the base stations. [00881 As used herein, "cell" refers to the area of signal coverage a base station provides, within which it is intended that a user equipment can establish a wireless radio connection to the base station. The cell is typically formed as a sector emanating from the antenna at the base station, with the nominal radius of the coverage area depending on a range of intrinsic factors including the frequency band, the Radio Access Technology, and antenna and base station configuration and power, and on extrinsic factors such as the topology, man-made structures and atmospheric conditions. The nominal coverage provided by a cell may differ from the location and distance distribution of user equipment actually establishing a connection with the base station. For example, in an urban environment the density of users may be such that a large number of user equipment are densely located near the base station (e.g. within 1-2 km) such that the majority of the connections are well within the nominal coverage area (which may be e.g. around 12 km for LTE 2100 MHz band). However, connections with user equipment beyond this nominal distance may be established if the user equipment receives a sufficiently strong signal from the base station. Generally, the further away from a base station, the lower the signal strength, the more unstable the signal and the lower the bandwidth. As such the “coverage” a base station can provide within different locations throughout a cell in terms of the number and stability of wireless connections and the data and voice throughput that can be supported thereby depends on a number of local factors including user demand, topology, the built environment, weather, etc, and is thus subject to change.

[0089] As used herein, "user equipment" refers to any suitable mobile or stationary device that can transmit and receive wireless communication signals and can form wireless connections with a base station of a wireless communication network, to thereby conduct voice and / or data communications through a core network of the wireless cellular network with other user equipment and nodes within the network, or with external networks such as the Internet. UEs can be embodied by any of a number of types of devices including but not limited to printed circuit (PC) cards, compact flash devices, external or internal modems, wireless phones, smartphones, tablets, tracking devices, asset tags, Internet of Things (loT) devices and so on.

[0090] FIG. 1 is a schematic illustration of an example of a wireless communication network 102. The wireless communication network 102 may comprise a mobile telecommunications network 102 which may also be referred to as a cellular telecommunications network 102. The network 102 includes a first base station 104, a second base station 106 and a core network 108. An electronic device 110 is shown as being attached to the network 102. The electronic device 110 may also be referred to as a user equipment (UE). [00911 The first base station 104 and the second base station 106 provide network coverage using a Radio Access Technology (RAT). Well known examples of RATs which may be employed in a mobile telecommunications network include, the Global System for Mobile Communications (GSM), the Universal Mobile Telecommunications System (UMTS), Long-Term Evolution (LTE) and 5G New Radio (NR). Each of these RATs are well-documented and subject to industry standardisation. The methods and apparatus disclosed herein are not however limited to the RATs listed above and may be implemented using advanced and future RATs such as, 5G advanced and sixth generation (6G) technologies.

[0092] Typically, each RAT is assigned one or more frequency bands over which radio communications may be transmitted and received using the respective RAT. Generally, different RATs are assigned different frequency bands so as to avoid frequency overlaps between different RATs. Communications using a given RAT therefore generally use one or more frequency bands assigned to the given RAT and use communication protocols adhering to the relevant standards for the given RAT. The first base station 104 and the second base station 106 may use the same RAT, or different RATs. Each of the first and second base stations may comprise a plurality of nodes.

[0093] The core network 108 may comprise any suitable network infrastructure for providing access to one or more network services, as is commonly known in the art. For example, the core network 108 may comprise an Evolved Packet Core (EPC). Additionally or alternatively, the core network 108 may comprise a Next Generation Core (NextGen Core) or 5G Core (5GC). Additionally or alternatively, the core network 108 may comprise a core network for use with 5G advanced, 6G and / or other future network technology.

[0094] In the example depicted in FIG. 1, the first base station 104 and the second base station 106 are connected to a common core network 108. That is, the first base station 104 and the second base station 106 may provide connectivity to the same core network 108. In the example shown in FIG. 1, a connection 112 is provided between the first base station 104 and the core network 108. Furthermore, a connection 114 is provided between the second base station 106 and the core network 108. In addition, a connection 116 may also be provided between the first base station 104 and the second base station 106. The connections 112-116 may be provided using one or more nodes provided at each end of the connection.

[0095] As is shown in FIG. 1, a first geographical coverage area 118 over which the first base station 104 provides network coverage and a second geographical coverage area 120 over which the second base station 106 provides network coverage may at least partially overlap. The geographical coverage areas 118, 120 may be referred to as cells. [00961 The at least partial overlap between the coverage areas 118, 120 provided by the first 104 and second 106 base stations may enable UE 110 situated in a region of coverage overlap to connect to the core network 108 through either or both of the first 104 and the second 106 base stations. The first base station 104 may exchange communications, such as radio frequency signals with a UE 110 over a first air interface 122. The second base station 106 may exchange communications with a UE 110 over a second air interface 124. Exchanging communications may comprise transmitting and / or receiving communications in one or more frequency bands assigned to a RAT used by the base station and using communication protocols specified for the RAT.

[0097] The network 102 illustrated in FIG. 1 and described above, is provided merely as an illustrative example of components and connections which may form all or part of a wireless communication network communication network. It will be appreciated that in practice a wireless communication network may include many more components than those shown in FIG. 1 and may provide wireless connectivity to many more devices than the single UE 110 shown in FIG. 1. For example, a wireless communication network may comprise a plurality of base stations (which may comprise more than two base stations) arranged over a wide geographic area to provide network coverage to UEs 110 over a wide geographic area.

[0098] In order to provide network connectivity to a UE 110, the network 102 allocates one or more radio parameters to the UE 110 for use when attached to and communicating with the network 102 and using an air interface 122, 124. Radio parameters may be allocated to UEs 110 which attach to the network in an effort to make an efficient use of resources in the network 102, to provide a desired quality of service to UEs 110 attached to the network 102 and / or to reduce radio interference in the network.

[0099] References herein to radio parameters is intended to refer to parameters which have a plurality of settings which can be allocated. For example, a first setting or a second setting may be allocated for a first radio parameter. Allocation of a radio parameter for a UE may comprise determining a setting for the radio parameter to be used for the UE. The terminology allocating a radio parameter and determining a setting for the radio parameter may be used interchangeable herein to refer to the same process.

[0100] The allocation of radio parameters may be carried out by a base station 104, 106 and for UEs attached to the base station 104, 106. More generally, the allocation of radio parameters may be carried out by a Radio Access Network (RAN) node. A RAN node as referred to herein may comprise any network node or function which is configured to perform all or part of the functionality of providing all or part of a RAN. For example, a RAN node may comprise all or part of a base station such as a gNB. In some examples, the functionality of a base station may be split across more than one RAN node. For example, the functionality of a base station may be provided by at least one Distributed Unit (DU or gNB-DU) and at least one Centralized Unit (CU or gNB-CU). A gNB-CU may for example, control a plurality of gNB-DU. In such examples, radio parameter allocation may be performed by a gNB-DU and / or a gNB-CU.

[0101] In at least some examples, radio parameters may be separately allocated for each UE attached to a RAN node (e.g., different devices attached to the same RAN node may be allocated different radio parameters) and / or may be allocated for a plurality of UEs attached to a RAN node (e.g., the same radio parameter may be allocated to a plurality of devices attached to the same RAN node).

[0102] Radio parameters which are allocated to a UE 110 may include at least some radio parameters which are allocated on different time scales. For example, reference is made herein to semi-static radio parameters and dynamic radio parameters. In general, settings are determined for dynamic radio parameters for a UE more frequently than settings are determined for semi-static radio parameters. That is, a time period for which a setting of a dynamic radio parameter is used for a given UE 110 (before a new setting of the dynamic radio parameter is determined) is generally shorter than a time period for which a setting of a semi-static radio parameter is used for the given UE 110 (before a new setting of the semistatic radio parameters is determined). To put it another way, a time period in between successive allocation of radio parameters for a given UE 110 may be less for dynamic radio parameters than semi-static radio parameters.

[0103] Dynamic radio parameters may comprise radio parameters which a setting is determined for each instance of downlink and / or uplink of data. For example, dynamic radio parameters may allocated for each data transmission to a UE 110 (downlink data) and / or for each data transmission from a UE 110 (uplink data). Dynamic radio parameters may allocated for transmission of each transport block. Dynamic radio parameters may allocated for use during one or a few (e.g., less than 10) transmission time interval(s) (TTI). Dynamic radio parameters may therefore be allocated for use during a time period of the order of milliseconds (e.g., a few tens of miliseconds). Dynamic radio parameters may be allocated for use during a time period which is less than about 500 milliseconds (ms). For example, dynamic radio parameters may be allocated for use during a time period which is less than about 100 ms.

[0104] Dynamic radio parameters may be communicated to a UE 110 (e.g., from a RAN node) using downlink control information (DCI) and / or uplink control information (UCI). Examples of dynamic radio parameters may include (but are not necessarily be limited to) radio resources for use to exchange data, a transmission power for use to exchange data and / or modulation and / or coding used for data transmissions.

[0105] Semi-static radio parameters may comprise radio parameters which are allocated for use during a plurality of instances of downlink and / or uplink of data. That is, semi-static radio parameters may not be allocated for each instance of downlink and / or uplink of data. Semi-static radio parameter may be allocated for use for transmission of a plurality of transport blocks. Semi-static radio parameters may be allocated for use over a plurality of TTIs (for example, tens of TTIs, several tens of TTIs or even longer). Semi-static radio parameters may, for example, be allocated for use during a time period which is greater than a few hundred milliseconds (for example, greater than about 500 milliseconds).

[0106] In at least some examples, semi-static radio parameters may be communicated to a UE 110 using Radio Resource Control (RRC) messages. RRC messages are messages which utilise the RRC protocol. RRC is a protocol used in standardised RATs (including UMTS, LTE, 5G and NR) to establish, release, reconfigure and maintain radio communication between a RAN node and a UE 110. Examples of situations in which RRC messages may be exchanged between a UE 110 and a RAN node may include (but are not necessarily limited to) when a UE 110 first attaches to a RAN node (connection establishment), when a UE 110 detaches from a RAN node (connection release), when a radio bearer is established, reconfigured and / or released, when connection mobility procedures are performed (e.g., cell handover or context transfer, cell selection and / or re-selection and / or inter-RAT mobility), paging a UE 110 in an RRC inactive or disconnected state and / or outer loop power control.

[0107] In general RRC messages are typically exchanged and RRC procedures are typically used during establishment, reconfiguration and / or release of a connection and / or bearer (e.g., a signalling radio bearer (SRB) and / or a data radio bearer (DRB)). In contrast to, for example, exchange of DCI and / or UCI, RRC messages are not typically exchanged for each instance of data transmission and / or reception (e.g., for each transport block). During typical operation of a UE 110, RRC procedures are carried out relatively infrequently (when compared, for example, to other forms of control signalling). Correspondingly, radio parameters which are allocated and communicated to a UE 110 using RRC messages may be considered to be semistatic radio parameters.

[0108] Semi-static radio parameters may comprise at least one radio parameter which configures a procedure to be used by a UE 110 and / or a base station when exchanging data (uplink data and / or downlink data). Examples of semi-static radio parameters may include (but are not necessarily limited to) at least one parameter related to the configuration of a Hybrid Automatic Repeat Request (HARQ) procedure, at least one parameter related to the configuration of a discontinuous reception (DRX) procedure, at least one parameter related to the configuration of a discontinuous transmission (DTX) procedure, at least one parameter related to the configuration of a procedure carried out using the Medium Access Control (MAC) protocol, at least one parameter related to the configuration of a procedure carried out using the Radio Link Control (RLC) protocol, and / or at least one parameter related to the configuration of a procedure carried out using the Packet Data Convergence Protocol (PDCP).

[0109] Examples of parameters related to the configuration of a HARQ procedure may include (but are not necessarily limited to) a maximum number of HARQ processes to be carried out for each data transmission.

[0110] Examples of parameters related to the configuration of a DRX may include (but are not necessarily limited to) a duration of a DRX cycle (e.g., DRX cycle), a duration of “ON” time within one DRX cycle (e.g. OnDurationTimer), a duration of time that a UE 110 should remain “ON” after reception of a Physical Downlink Control Channel (PDCCH) message (e.g., drx-Inactivity timer), a maximum number of consecutive PDCCH subframes that a UE 110 should remain active to wait for an incoming re-transmission (e.g., drx-Rctransmission tinier), a parameter related to a short DRX cycle which can be implemented (e.g., shortDRX cycle) and / or a duration of time that a UE 110 should use a short DRX cycle (e.g., drxShortCycleTimer).

[0111] Examples of parameters related to the configuration of a DTX may include (but are not necessarily limited to) one or more parameters which configure the timing and / or procedures to be used in a DTX cycle.

[0112] Examples of parameters related to the configuration of a procedure carried out using the MAC protocol may include (but are not necessarily limited to) one or more parameters used to configure MAC parameters for a cell group. Such parameters may, for example, include one or more parameters set in MAC-CellGroupConfig, as defined for 5G in the 3GPP Standards Specification TS 38.331 which is incorporated herein by reference in its entirety. For example, parameters related to the configuration of a procedure carried out using the MAC protocol may include one or more of a parameter which sets the use of a pre-emptive Buffer Status Report (BSR), a parameter which limits the sending of a Channel State Information (CS1) report to an on-duration period of a DRX cycle, a timer after which a connection is released if there is data inactivity for duration of the timer, a parameter for configuring a DRX procedure, a parameter which sets a prioritisation of a MAC entity for overlapping grants and scheduling requests, a parameter which indicates a scheduling request configuration applicable for a beam failure recovery on a secondary cell, a parameter which indicates a scheduling request configuration applicable for consistent uplink Listen-Before-Talk (LBT) recovery on a secondary cell, a parameter which determines skipping of uplink transmissions, and / or a parameter used to configure a time-alignment group.

[0113] Examples of parameters related to the configuration of a procedure carried out using the RLC protocol may comprise one or more parameters for setting an RLC mode and / or for configuring a radio bearer such as a SRB and / or a DRB. Examples of parameter related to the configuration of a procedure carried out using the RLC protocol may include (but are not necessarily limited to) one or more parameters set in RLC-config as defined for 5G in the 3GPP Standards Specification TS 38.331. For example, parameters related to the configuration of a procedure carried out using the RLC protocol may include one or more of a parameter indicated whether to operate using an RLC transparent mode (TM), acknowledged mode (AM) or unacknowledged mode (UM), a parameter which sets a maximum number of transmissions for use in RLC acknowledged mode, a number of bytes for which polling of peers should be performed in RLC acknowledged mode, a number of packet data units (PDU) for which polling of peers should be performed in RLC acknowledged mode, a sequence number field size, a timer for a transmitting side for retransmitting a poll in RLC acknowledged mode, a timer for a receiving side to detect loss of RLC PDUs at lower layers in RLC acknowledged mode, and / or a timer for a receiving side to prohibit transmission of a STATUS PDU in RLC acknowledged mode.

[0114] Examples of parameters related to the configuration of a procedure carried out using the PDCP protocol may comprise one or more configurable parameters for SRBS and / or DRBs. Examples of parameters related to the configuration of a procedure carried out using the PDCP protocol may include (but are not necessarily limited to) one or more parameters set in PDCP-config as defined for 5G in the 3GPP Standards Specification TS 38.331. For example, parameters related to the configuration of a procedure carried out using the PDCP protocol may include one or more of a parameter for disabling ciphering, a timer for discarding a service data unit (SDU) and a corresponding PDU at a transmitting PDCP entity, a parameter indicated whether a PDCP entity continues or resets a robust header compression (ROHC) protocol during PDCP re-establishment, a parameter indicating a PDCP uplink duplication state for an associated RLC entity, a parameter configuring ethemet header compression, a parameter configuring ROHC, a parameter indicating whether or not integrity protection is configured for a radio bearer, a parameter indicating a maximum value for a Context Identifier (CID), a parameter which configures uplink data transmission when more than one RLC entity is associated with the PDCP entity, a parameter which sets whether a PDCP SDU is delivered to upper layers after performing header decompression, a parameter which indicates whether or not uplink duplication status is configured and activated, a PDCP sequence number size for downlink, a PDCP sequence number size for uplink, a parameter which indicates a cell group identification and a logical channel identification (LCID) of a primary RLC entity, a parameter indicating an LCID of a split secondary RLC entity, a parameter indicating whether a data radio bearer is configured to send a PDCP status report in the uplink, and / or a timer used to detect loss of PDCP data PDUs, a threshold amount of data for transmitting data to a primary and secondary RLC entity if a split secondary RLC entity is configured. [01151 Radio parameters, such as semi-static radio parameters, may be allocated to different UEs attached to a network 102 with a view to efficiently utilising the resources available in the network whilst reducing or minimising interference. Additionally or alternatively radio parameters, such as semi-static radio parameters may be allocated so as to seek to provide a minimum Quality of Service (QoS) to one or more UEs attached to the network. For example, radio parameters may be allocated so as to prioritise a UE and / or network traffic which has a higher QoS requirement associated with it.

[0116] Whilst a desired or minimum QoS may be accounted for when allocating radio parameters, QoS measures are concerned with parameters such as data rate, error rate and delay, which do not directly translate into a quality which is experienced by a user of a UE. Methods and apparatus are described herein in which measures of a quality of a user experience which is provided by a UE whilst providing a service and using a wireless communication network are taken into account when allocated radio parameters, such as semistatic radio parameters.

[0117] Measures which are indicative of a quality of user experience may be referred to herein as Quality of Experience (QoE) metrics or data. QoE metrics or data as referred to herein directly relate to a quality of a user experience and may be linked to a specific end user service or end user service type (e.g., playout and / or streaming of media content). QoE metrics or data may, for example, relate to factors which are directly observable by a user whilst receiving an end user service (e.g., delay, quality, etc. experienced when receiving a service such as playout of media content). QoE metrics or data is typically measured and / or collected at an application layer (e.g., by an application providing an end user service). QoE metrics or data differs from other measures of network performance such as, for example, radio signal measurements and / or Quality of Service (QoS) measures which are not typically measured and directly experienced at an application layer but are determined at a radio layer and relate to underlying measures of radio performance.

[0118] QoE measures or data may comprise one or more metrics such as an average throughput, an initial playout delay, a buffer level, a play list, a playout delay for media startup, jitter, an intermediate playout delay, a perceptual quality and device information. QoE measures or data may relate to a quality of a user's experience whilst a streaming of media content over the wireless communication network. QoE measures or data may be collected during a QoE profile time period which may, for example, comprise a few tens of milliseconds.

[0119] An average throughput may comprise an average throughput which is measured by a UE during a given measurement time interval. An average throughput may be measured whilst a UE provides a service (such as streaming or playout of media content and extended realtiy (XR) experience etc.) using the network.

[0120] An initial playout delay may comprise a measure of an initial delay at the start or providing a service using the network. For example, an initial playout delay may comprise a measure of a time period (e.g., in miliseconds) of an initial delay at the start of streaming of media content (such as video content) using the network.

[0121] A buffer level may comprise a measure of the occupancy level of a buffer whilst providing a service using the network. For example, a buffer level may comprise a measure of the occupancy level of a buffer during streaming or playout of media content using the network.

[0122] A play list may comprise a list of time periods during which samples of a single representation were continuously rendered during playout of media content and one or more properties associated with each time period. The one or more more associated properties may, for example, comprise a timestamp of a user action that starts the time period, a presentation time at which playout was requested by the user action, a type of user action which triggered playout (e.g., new playout request, resume from pause, other user request, etc.) and / or a list of periods of continuous rendering of content with associated parameters such as a representation identification, timing information, playback speed and reasons for stopping continuous rendering.

[0123] Device information which may be included in QoE metrics or data may comprise information about a displayed video resolution and / or physical screen characteristics such as a width and height of the screen (in milimiteres and / or in number of pixels). [01241 More information and definitions regarding QoE metrics is provided, for example, in the 3GPP Standards Specification TS 26.247 which is incorporated herein by reference in its entirety.

[0125] FIG. 2 is a flowchart of an example method 200 of allocating radio parameters for a UE. As will be explained in further detail below, the method of allocating radio parameters utilises data indicative of a QoE being provided at one or more UEs in the network 102 so as to allocated radio parameters in dependence on QoE data. The method 200 of FIG. 2 may be implemented using one or more computing devices. For example, the method 200 of FIG. 2 may be implemented at least in part by a RAN node (e.g., base station).

[0126] At step 202 radio parameter data indicative of a first setting of at least one radio parameter is received. The first setting of the at least one radio parameter comprises a setting which is being used for a first UE 110 in the network 102, whilst the first UE 110 provides a service using the network 102. The radio parameter data may be received from another device (e.g., the first UE 110) or may be received from a function or component of the same device which implements the method 200. In examples, in which the method 200 of FIG. 2 is implemented by a RAN node, the RAN node may already be aware of the first setting of the at least one radio parameter (e.g., the RAN node may have allocated the first setting). In such examples, the radio parameter data may be received by retrieving it internally (e.g., from memory and / or being passed from another function or routine).

[0127] The at least one radio parameter may comprise a semi-static radio parameter. For example, the at least one radio parameter may comprise one or more of a parameter related to the configuration of a HARQ procedure, a parameter related to the configuration of a DRX procedure, a parameter related to the configuration of a DTX procedure, a parameter related to the configuration of a procedure carried out using a MAC protocol, a parameter related to the configuration of a procedure carried out using a RLC protocol, and / or a parameter related to the configuration of a procedure carried out using a PDCP protocol.

[0128] The at least one radio parameter may comprise a plurality of radio parameters. In such examples, the first setting may comprise settings for each of a plurality of radio parameters.

[0129] At step 204 QoE data is received from the first UE 110. The QoE data is indicative of a quality of a user experience provided by the first UE 110 whilst the first UE 110 provides the service using the first setting of the at least one radio parameter. The QoE data may be indicative of a quality of experience over a period of time. For example, QoE metrics may be collected over a profile period of time over which a QoE profile is evaluated. The profile period of time may be greater than 1 millisecond and may, for example, be a period of a few tens of milliseconds. QoE data may therefore be indicative of a quality of experience which has occurred over a profile period of time. The QoE data may relate to a quality of a user experience which is linked to a specific end user service or end user service type (e.g., playout and / or streaming of media content). The QoE data may be collected by an application layer of the first UE 110 (e.g., by an end user application). The QoE data may comprise one or more metrics such as an average throughput, an initial playout delay, a buffer level, a playlist, a playout delay for media start-up, a perceptual quality, jitter, an intermediate playout delay and device information. The QoE data may relate to a quality of a user's experience whilst streaming of media content using the wireless communication network. [01301 Whilst not shown in FIG. 2, the method 200 of FIG. 2 may further comprise receiving radio coverage data. The received radio coverage data may comprise and / or may be based on one or more measurements made by the first UE 110. The radio coverage data may, for example, comprise one or more of a measure of a received signal power (e.g. a received reference signal power (RSRP)), a received signal quality (e.g. a received reference signal quality (RSRQ)), a signal to noise ratio (e.g. a signal to interference and noise ratio (SINR)), and / or a channel quality indicator (e.g. channel state information (CSI)). The radio coverage data may be associated with a serving cell and / or one or more neighbouring cells.

[0131] Steps 202 and 204 have been described in the context of receiving data (radio parameter data, QoE data and optionally radio coverage data) from a first UE 110. In some examples, the method 200 may further comprise receiving data (such as radio parameter data, QoE data and / or radio coverage data) from one or more other UEs attached to the network 102. For example, the data may be received from a plurality of UEs attached to the network 102, which may include UEs attached to the same RAN node as the first UE 110 and / or may include UEs attached to a different RAN node (e.g., a neighbouring RAN node) as the first UE 110

[0132] At step 206 the received radio parameter data and the received QoE data are provided to a parameter allocation modelter parameter allocation model. The parameter allocation may be implemented by a RAN node (e.g., base station) and / or may be implemented by another component such as an Operations, Administration and Maintenance (0AM) function. The parameter allocation model is configured to determine, in dependence on the received radio parameter data and QoE data a second setting of the at least one radio parameter. The second setting of the at least one radio parameter is a setting for use by the first UE 110 to receive at least a prescribed QoE whilst using the second setting of the at least one radio parameter. [01331 As is described in further detail elsewhere herein, the parameter allocation model may be configured through training. For example, the parameter allocation model may be configured through training based on training data records comprising data related to settings of the at least one radio parameter which have been used in the network and QoE data which indicates a QoE which resulted from the use of the setting of the at least one radio parameter.

[0134] The parameter allocation model may comprise an artificial intelligence and / or machine learning model. In some examples, the parameter allocation model may comprise a predictive model configured through training to determine a predicted QoE which is expected to result from an input setting of the at least one radio parameter. The predictive model may, for example, be trained using a supervised learning process and based on a plurality of training data records.

[0135] The prescribed QoE may represent a minimum QoE which is desired or required for the first UE 110. The prescribed QoE may comprise a pre-defined measure of QoE. The prescribed QoE may be dynamic. For example, the prescribed QoE may be updated dependent on an application, service or task being performed or provided by the first UE. The prescribed QoE may comprise a value of at least one metric representative of a QoE. For example, the prescribed QoE may comprise a value for at least one metric including one or more of an average throughput, an initial playout delay, a buffer level, a playlist, a playout delay for media start-up, a perceptual quality, jitter and / or intermediate playout delay. It will be appreciated that an improvement in a QoE may correspond with an increase in some metrics of QoE (such as an average throughput) but may correspond with a decrease in other metrics of QoE (such as a buffer level and / or playout delay). The prescribed QoE may comprise a minimum of a first metric for which an increase represents an improvement (such as an average throughput) and / or may comprise a maximum of a second metric for which a decrease represents an improvement (such as a buffer level and / or playout delay).

[0136] In some examples, the parameter allocation model may be configured to determine the second setting of the at least one radio parameter for use for the first UE 110 for the first UE 110 to receive an improved QoE whilst using the second setting.

[0137] As was described above, in some examples data (e.g., radio parameter data, QoE data and / or radio coverage data) may be received from one or more other UEs. Data received from other UEs may also be provided to the parameter allocation model and the determination made by the parameter allocation model may be further based on the data received from the other UEs. In such examples, the parameter allocation model may be configured to determine the second setting of the at least one radio parameter so as to provide a prescribed QoE both for the first UE and for one or more other UEs (e.g., other UEs for which data is provided to the parameter allocation model).

[0138] In some examples, the parameter allocation model may be configured to allocate settings of the at least radio parameters for use by a plurality of UEs. For example, the parameter allocation model may be configured to determine optimum or improved settings of the at least one radio parameter for a group of UEs. In some examples, the parameter allocation model may be configured to determine a plurality of second settings of the at least one radio parameter, where the plurality of second settings comprise a setting for each of a plurality of UEs. For example, the parameter allocation model may determine a second setting of the at least one radio parameter for use by the first UE and second settings of the at least one radio parameter for use by each of the one or more second UEs.

[0139] At step 208 the determined second setting of the at least one radio parameter determined by the parameter allocation model is sent to the first UE 110. The second setting may, for example, be sent to the first UE 110 as part of a control message such as an RRC message. The message may be configured to cause the first UE 110 to use the second setting of the at least one radio parameter for data transmission and / or reception over the network 102.

[0140] As was explained above, in some examples the parameter allocation model may be configured to determine a plurality of second settings of the at least one radio parameter, where the plurality of second settings comprise a setting for each of a plurality of UEs. In such examples, each of the plurality of second settings may be sent to the corresponding UEs.

[0141] As was described above, in at least some examples the parameter allocation model may comprise a predictive model. For example, a predictive model may be configured through training to determine a predicted QoE which is expected to result from an input setting of at least one radio parameter. Such a predictive model may, for example be configured through training based on training data records which may, for example, comprise historical settings of at least one radio parameter and QoE data collected during use of the historical settings. A predictive model may, for example, be configured through training using a supervised learning process.

[0142] FIG. 3 is a flowchart of an example method 302 of the operation of a radio parameter allocation model which comprises a predictive model. All or part of the method 302 of FIG. 3 may be carried out by any suitable computing device (or combination of computing devices) which implements a parameter allocation model as described herein. For example, all or part of the method 302 may be implemented by a RAN node (e.g., a base station). The method of FIG. 3 may form part of a method as described above with reference to FIG. 2. For example, the method of FIG. 3 may form all or part of step 206 of the method of FIG. 2.

[0143] At step 304, a first setting of at least one radio parameter and QoE data are received. The at least one radio parameter and / or the QoE data may have any of the features or properties of radio parameters and QoE data as described herein. The QoE data may be indicative of a QoE provided at the first UE whilst using the received setting of the QoE data. The received setting and the received QoE data may represent a setting of the at least one radio parameter which is currently being used and a QoE which has resulted from use of the setting.

[0144] At step 306 a new potential input setting of the at least one radio parameter is determined. The new potential input setting setting may comprise a new setting which the parameter allocation model determines may bring about an improvement in QoE.

[0145] At step 308 a predictive model is implemented to determine a predicted QoE which is expected to result from the new potential input setting. The predictive model may be provided with at least one input which may comprise at least the new potential input setting. The at least one input may further comprise one or more other inputs which may, for example, relate to the state of the network 102. For example, data indicating radio coverage, distribution, network load etc., for one or more UEs may be further provided as inputs to the predictive model. The predicted QoE which is determined by the predictive model may comprise a value of at least one metric representative of a QoE. For example, the predicted QoE may comprise a value for at least one metric including one or more of an average throughput, an initial playout delay, a buffer level, a play list, a playout delay for media startup, a perceptual quality, jitter and / or intermediate playout delay.

[0146] In some examples, the predictive model may determine a predicted QoE associated with a plurality of UEs 110 attached to the network 102. In such examples, the predictive model may be provided with a setting of the at least one radio parameter and QoE data (and / or one or more further inputs) associated with each of the UEs for which a predicted QoE is to be determined.

[0147] At step 310 it is determined whether or not the predicted QoE meets a criteria. The criteria may for example, comprise whether or not the predicted QoE is at least a prescribed QoE as was described above. Additionally or alternatively the criteria may comprise whether or not the predicted QoE represents an improvement in QoE (relative to a current QoE). Additionally or alternatively, the criteria may comprise whether or not the predicted QoE represents an optimum QoE. In some examples, steps 306 and 308 may be iterated through a plurality of times to determine a plurality of predicted QoEs for a plurality of new potential input settings. In such examples, step 310 may comprise selecting a predicted QoE from a plurality of predicted QoEs which represents an optimum QoE. In some examples, step 310 may comprise comparing a predicted QoE determined in a current iteration with predicted QoEs determined in previous iterations in order to determine whether the predicted QoE determined in the current iteration is an optimum QoE of the plurality of predicted QoEs.

[0148] If the criteria is met in step 310 then the method proceeds to step 312 in which the new potential input setting is determined as the second setting of the at least one radio parameter. That is, if the criteria is met in step 310 then the new potential input setting is adopted as the second setting of the at least one radio parameter.

[0149] If the criteria is not met in step 310 then the method returns to step 306 to determine a new potential input setting for which the predictive model is implemented in step 308. In this way, an iterative process may be used with a predictive model to determine the second setting of the at least one radio parameter.

[0150] The method of FIG. 3 is presented by way of illustrative example only and the parameter allocation model may operate according to alternative method steps to those described above with reference to FIG. 3. For example, as was described above, in some examples the parameter allocation model may operate according to a reinforcement learning process.

[0151] FIG. 4 is a flowchart of an example method 400 of training a radio parameter allocation model for determining a setting of at least one radio parameter for use for a UE. The parameter allocation model may comprise any parameter allocation model as described herein and may be suitable for use in any method of allocated radio parameters as described herein. For example, a parameter allocation model trained according to the method 400 of FIG. 4 may be suitable for use in a method described above with reference to FIG. 2 and / or FIG. 3.

[0152] The method of FIG. 4 may be implemented by any suitable computing device, or combination of computing devices. For example, the method 400 of FIG. 4 may be implemented by one or more of a RAN node and / or an 0AM.

[0153] At step 402 a plurality of training data records are received. Each training data record comprises at least an historical setting of at least one radio parameter used for a UE and QoE measurement data indicative of a quality of user experience provided by the UE whilst using the historical setting of the at least one radio parameter. Each training data record may be representative of a QoE provided by a UE and a setting of at least one radio parameter which was used whilst the QoE was provided.

[0154] The at least one radio parameter may comprise one or more of any of the radio parameters described elsewhere herein. For example, the at least one radio parameter may comprise a semi-static radio parameter as described herein. The at least one radio parameter may comprise one or more of a parameter related to the configuration of a HARQ procedure, a parameter related to the configuration of a DRX procedure, a parameter related to the configuration of a DTX procedure, a parameter related to the configuration of a procedure carried out using a MAC protocol, a parameter related to the configuration of a procedure carried out using a RLC protocol, and / or a parameter related to the configuration of a procedure carried out using a PDCP protocol. The at least one radio parameter may comprise a plurality of radio parameters. In such examples, the historical setting may comprise a setting for each of the plurality of radio parameters.

[0155] The QoE measurement data may comprise QoE data as described elsewhere herein. For example, the QoE measurement data may comprise a value for at least one metric including one or more of an average throughput, an initial playout delay, a buffer level, a play list, a playout delay for media start-up, a perceptual quality, jitter and / or intermediate playout delay. The QoE measurement data may be determined at an application layer in a UE (e.g., by an end-user application running on the UE).

[0156] In some examples, a training data record may further comprise one or more other measurements or data. For example, a training data record may further comprise radio coverage data. Radio coverage data may comprise any of the properties of radio coverage data described elsewhere herein. For example, radio coverage data may comprise one or more of a measure of a received signal power (e.g. a received reference signal power (RSRP)), a received signal quality (e.g. a received reference signal quality (RSRQ)), a signal to noise ratio (e.g. a signal to interference and noise ratio (SINR)), and / or a channel quality indicator (e.g. channel state information (CSI)). The radio coverage data may be associated with a serving cell and / or one or more neighbouring cells. Radio coverage data may comprise, or otherwise be based on, measurements made by a UE during which the historical setting of the at least one radio parameter was used and during which the QoE measurement data was collected.

[0157] Training data records may be received for a plurality of different UEs. In some examples, training data records may be received for a plurality of different UEs which were attached to the a plurality of different cells at the times to which training data records relate. The training data records may serve to provide historical evidence of a QoE provided in the network and the conditions and settings in place whilst the QoE was provided. The training data records may therefore be suitable for training a model which may learn patterns present in the training data in order to determine future settings for bringing about a prescribed QoE.

[0158] At step 404 a parameter allocation model is trained in dependence on the received training data records. The parameter allocation model is trained for determining a setting of the at least one radio parameter for use for a UE and for the UE to provide at least a prescribed QoE whilst using the determined setting.

[0159] The prescribed QoE may be a prescribed QoE as described elsewhere herein. For example, the prescribed QoE may represent a minimum QoE which is desired or required for the first UE. In some examples, the parameter allocation model may be trained to determine the setting of the at least one radio parameter for use for a first UE for the first UE to receive an improved QoE whilst using the determined setting. That is, the parameter allocation model may be trained to determine a setting of the at least one radio parameter so as to bring about an improvement in the QoE provided at the first UE.

[0160] The parameter allocation model may be trained to determine the setting of the at least one radio parameter for use for the UE to bring about an overall improvement in a QoE provided by a plurality of UEs attached to the wireless communication network. For example, the parmater allocation model may be trained to determine a setting of the at least one radio parameter in order to bring about an improvement in a QoE provided at a plurality of UEs which includes the UE for which the setting is determined.

[0161] The parameter allocation model may be trained using a machine learning method. For example, the parameter allocation model may be trained using a supervised learning and / or a reinforcement learning method.

[0162] In some examples, the parameter allocation model may comprise a predictive model. Training the parameter allocation model may comprise training a predictive model to determine a predicted QoE which is expected to result from an input setting of the at least one radio parameter. A predictive model may be trained using a supervised learning process based on the training data records. In some examples, training a predictive model may comprise training a predictive model to determine predicted QoEs which are expected to be provided at a plurality of UEs.

[0163] As was explained above, training the parameter allocation model may comprise training the parameter allocation model using a reinforcement learning method. The reinforcement learning method may comprise a deep reinforcement learning method. In examples, in which a reinforcement learning method is used, steps 402, 404 of the method 400 of FIG. 4 may be carried out for a plurality of iterations to both train and implement the parameter allocation model. For example, the training data records received at step 402 may represent a current state of the network which may have resulted from an action (e.g., a setting of the at least one radio parameter) generated by a previous iteration of a reinforcement learning method. The training of the parameter allocation model at step 404 may comprise assessing the change of state (e.g., change in QoE) which resulted in the previous action (e.g., historical settings of the at least one radio parameter) and updating the parameter allocation model based on the change of state. The parameter allocation model may then be implemented to generate a new action in the form of a new setting for the at least one radio parameter. The method steps may then be repeated for a plurality of iterations.

[0164] A reinforcement learning method may comprise a multi-agent reinforcement learning (MARL) method. In such examples, different UEs attached to the network may have different agents associated with them. That is, each agent in a MARL process may be associated with a different UE. A MARL method may be carried out for a plurality of UEs so as to allocate radio parameters to each of the plurality of UEs as part of the same MARL method.

[0165] As was explained above, a reinforcement learning method generally takes a series of actions and observes changes in states which result from the actions. Through such a process, a reward which is associated with different actions is learned in order to choose future actions which maximise or optimise the reward. When applied to examples disclosed herein the actions in a reinforcement learning process may comprise an allocation of a setting for the at least one radio parameter. A state which is observed in a reinforcement learning process may comprise received QoE data which is associated with the first setting (a given action) of the at least one radio parameter. A reward in a reinforcement learning process may comprise an improvement in at least one metric of a QoE for one or multiple UEs. When applied to examples disclosed herein, in each iteration of a reinforcement learning process an action in the form of a second setting of the at least one radio parameter may be determined in order to seek a reward in the form of an improvement in at least one metric of a QoE for one or multiple UEs.

[0166] In some examples, training the parameter allocation model may comprise a centralized training for decentralized execution. In such examples, the parameter allocation model may be trained using data collected for a plurality of UEs in the network but may be executed independently for different UEs. In some examples, centralized or co-operative learning may be used. For example, a plurality of RAN nodes may provide data corresponding to UEs attached to each RAN node in order to perform a centralized learning 38 process. In some examples a co-operative learning process may comprise co-operation between different RAN nodes.

[0167] FIG. 5 is a signal flow diagram of an example method of allocating radio parameters for a user equipment. The method of FIG. 5 is presented by way of illustrative example only. Steps 510-522 of the method of FIG. 5 may be considered to form all or part of a method for training a parameter allocation model. Steps 526-532 of the method of FIG. 5 may be considered to form all or part of a method for allocating radio parameters for a UE.

[0168] The method illustrated in FIG. 5 is implemented using a UE 502, a first RAN node 504, a second RAN node 506, and an Operations, Administration and Maintenance (0AM) function 508. The UE 502 is a UE for which at least one radio parameter is to be allocated (i.e., a setting of the at least one radio parameter is to be determined for use). The UE 502 is attached to the first RAN node 504. For example, a cell operated by the first RAN node 504 may be a serving cell for the UE 502. The second RAN node 506 may comprise a neighbouring RAN node to the first RAN node 504. For example, the first RAN node 504 and the second RAN node 506 may be situated in a similar geographic vicinity. The first RAN node 504 and the second RAN node 506 may operate cells which neighbour each other and which may at least partially overlap. The first RAN node 504 and the second RAN node 506 may be communicatively coupled to each other (either via a direct communication interface or indirectly through one or more other entities). The 0AM 508 may for example be implemented as part of a core network 108.

[0169] At step 510 at least one radio parameter is allocated to the UE 502. That is, a setting of at least one radio parameter is determined for the UE 502 and sent to the UE 502. The allocation of at least one radio parameter at step 510 may be carried out according to any example method described herein or may be carried out according to any other allocation method (including those known from the prior art). The determined setting for the at least one radio parameter may be sent to the UE 502 as part of an RRC message. Following the allocation of the at least one radio parameter at step 510, the UE 502 may adopt the allocated at least one radio parameter for use when exchanging data using the wireless communication network. For example, in an example in which the at least one radio parameter comprises a parameter related to the configuration of a DRX procedure, the UE 502 may receive a new setting for the configuration of the DRX procedure at step 510 and may then adopt the new setting for the configuration of the DRX procedure for performing a DRX procedure.

[0170] At step 512, the UE 502 performs a measurement. The measurement made at step 512 may, for example, comprise a measurement of one or more metrics of a QoE provided at the UE 502 whilst providing a service using the at least one radio parameter which was allocated at step 510. For example, the measurement may comprise a measurement of one or more of an average throughput, an initial playout delay, a buffer level, a play list, a playout delay for media start-up, a perceptual quality, jitter and / or intermediate playout delay, whilst the UE 502 provides a service using the at least one radio parameter which was allocated in step 510.

[0171] The measurement may further comprise a measurement of one or more other parameters. For example, the measurement may comprise a measurement indicative of a radio coverage experienced at the UE 502 (which may be referred to as radio coverage data). Such measurements may comprise one or more of a measure of a received signal power (e.g. a received reference signal power (RSRP)), a received signal quality (e.g. a received reference signal quality (RSRQ)), a signal to noise ratio (e.g. a signal to interference and noise ratio (SINR)), and / or a channel quality indicator (e.g. channel state information (CSI)). Such measurements may be made for a serving cell (e.g., a cell provided by the first RAN node 504) and / or one or more neighbouring cells (e.g., a cell provided by the second RAN node 506). For example, measured radio coverage data may comprise one or more of a measure of a received signal power (e.g. RSRP), a received signal quality (e.g. RSRQ), a signal to noise ratio (e.g. SINR), and / or a channel quality indicator (e.g. CSI) associated with a serving cell, and may additionally or alternatively comprise one or more of a measure of a received signal power (e.g. RSRP), a received signal quality (e.g. RSRQ), a signal to noise ratio (e.g. SINR), and / or a channel quality indicator (e.g. CSI) associated with a serving cell associated with one or more neighbouring cells. Similarly to QoE measurements, measurements of radio coverage may be made for a time during which the at least one radio parameter which was allocated in step 510 was used by the UE 502.

[0172] At step 514 a measurement report is sent from the UE 502 to the first RAN node 504. The measurement report may comprise data indicative of measurements made at step 512. The measurement report may optionally further comprise one or more settings used by the UE 502 during a time period in which the measurements were made. For example, the measurement report may comprise a setting of at least one radio parameter which was used during the time period in which measurements were made. However, in some examples settings used by the UE 502 may have been determined or allocated by the first RAN node 504 (e.g., may have been sent to the UE 502 from the first RAN node 504 in step 510) and may therefore already be known to the first RAN node 504. In such examples, the measurement report need not necessarily include any settings used by the UE 502. [01731 At step 516 training data is sent from the first RAN node 504 to the 0AM 508. The training data may be based at least in part on the measurement report received at the first RAN node 504 in step 514. In particular, the training data may comprise at least a setting of at least one radio parameter which was used by the UE 502 and corresponding QoE data indicative of a QoE which was provided by the UE 502 whilst using the setting of the at least one radio parameter. As was described above, the setting of the at least one radio parameter may have been included in the measurement report sent to the first RAN node 504 in step 514 and / or may be determined by the first RAN node 504 according to a setting which was determined by the first RAN node 504 in step 510. The training data may further comprise one or more other properties or measurements such as radio coverage data. [01741 The training data may comprise a plurality of training data records, where each training data record comprises at least a setting of at least one radio parameter which was used by a UE and corresponding QoE data indicative of a QoE which was provided by the UE whilst using the setting of the at least one radio parameter. Training data records may further comprise one or more other properties or measurements such as radio coverage data. In some examples, steps 510, 512 and 514 may be carried out a plurality of times to generate a plurality of training data records corresponding to different instances of radio parameter allocation and corresponding QoE data.

[0175] The training data may relate to a plurality of UEs. For example, steps corresponding to steps 510, 512 and 514 may be performed for a plurality of UEs attached to the first RAN node 504. In this way training data records may be compiled which are representative of a QoE provided at a plurality of UEs and settings of at least one radio parameter which were used for the plurality of UEs and which resulted in the measured QoE.

[0176] At step 518 training data is sent from the second RAN node 506 to the 0AM 508. Whilst not shown in the signal flow diagram of FIG. 5, steps corresponding to steps 510, 512 and 514 may be carried out for one or more UEs attached to the second RAN node 506. In this way the second RAN node 506 may compile training data records which are representative of a QoE provided at one or more UEs attached to the second RAN node 506 and settings of at least one radio parameter which were used for the one or more UEs and which resulted in the measured QoE. Whilst not shown in FIG. 5, training data may additionally or alternatively be sent to the 0AM 508 from one or more further RAN nodes in the network.

[0177] At step 520 a training process is performed at the 0AM 508. The training may comprise training of a parameter allocation model as described above. The training is carried out using the training data received at the OAM 508. For example, the training data sent in step 516 and / or the training data sent in step 518 may be used in the training process carried out at step 520. The training process may be carried out using any suitable training method. For example, the training process may comprise executing a machine learning algorithm. For example, the training process may comprise supervised learning. Additionally or alternatively, the training process may comprise reinforcement learning.

[0178] In some examples, different instances of a parameter allocation model may be trained for different geographical areas, different time periods (e.g., different times of day such as day and night), and different conditions (e.g., weather conditions and / or network load conditions). Each instance of a parameter allocation model may be trained using training data which corresponds to the geographical area, time period and / or condition to which the parameter allocation model corresponds.

[0179] At step 522, a trained parameter allocation model is deployed by sending the trained parameter allocation model to the first RAN node 504. The trained parameter allocation model may comprise configuration parameters (e.g., weights) of the parameter allocation model which were determined during the training process (e.g., which were determined to best fit the training data).

[0180] At step 524 a further measurement is carried out at the UE 502. The measurement carried out at step 524 may correspond to the measurement taken at step 512 and may include any of the properties or features described above with reference to step 512. For example, the measurement may comprise a measurement of QoE data indicative of a QoE provided at the UE 502 and may further comprise one or more further measurements such as radio coverage data.

[0181] At step 526 a measurement report is sent from the UE 502 to the first RAN node 504. The measurement report sent at step 526 may correspond to the measurement report sent at step 514 and may include any of the properties or features described above with reference to step 514 (except that the measurement report sent at step 526 relates to measurements made at step 524, whereas the measurement report sent at step 514 relates to measurements made at step 512). The measurement report may, for example, comprise QoE data based on measurements made at step 524. The measurement report may further comprise radio coverage data based on measurements made at step 524.

[0182] Whilst not shown in FIG. 5, the method may further comprise receiving measurement reports from one or more other UEs attached to the first RAN node 504. [01831 At step 528 input data may be received at the first RAN node 504 and from the second RAN node 506. The input data may comprise (or may be based on) one or more measurement reports received at the second RAN node 506 and from one or more UEs attached to the second RAN node 506. For example, the input data may comprise QoE data indicative of a QoE provided at one or more UEs attached to the second RAN node 506 and may further comprise a setting of at least one radio parameter which was in use when the QoE data was collected. Whilst not shown in FIG. 5, in some examples input data may additionally be received from one or more further RAN nodes.

[0184] At step 530 model inference is carried out at the first RAN node 504. Model inference may comprise providing inputs to a trained parameter allocation model in order for the trained parameter allocation model to output a setting of the at least one radio parameter to be used for a UE. The trained parameter allocation model, may comprise a model configured through training, where the configuration was received at step 522 as described above. The inputs provided to the parameter allocation model may comprise data included in the measurement report received in step 526 and / or data included in the input data received at step 528 (and / or any further input data received at the first RAN node 504).

[0185] As was described in detail above, the parameter allocation model may output a setting of at least one radio parameter (which is described herein as the second setting) which is determined by the parameter allocation model. The determined setting of the at least one radio parameter may comprise a setting which is determined by the parameter allocation model to provide at least a prescribed QoE. The determined setting may comprise a setting to be used by the UE 502. In some examples, the parameter allocation model may comprise settings of at least one radio parameter to be used by a plurality of UEs (which may include the UE 502).

[0186] At step 532, radio parameter configuration is carried out. Radio parameter configuration may comprise sending a determined setting of at least one radio parameter to at least one UE. For example, the radio parameter configuration may comprise sending a determined setting of the at least one radio parameter (as determined in step 530) to at least one UE (which may include the UE 502). A determined setting of at least one radio parameter may be sent to a UE as part of an RRC message. The radio parameter configuration may further comprise, a UE and / or a RAN node adopting the determined setting of the at least one radio parameter when exchanging data over the wireless communication network.

[0187] Whilst not shown in FIG. 5, the method may further comprise sending feedback to the 0AM 508. For example, feedback relating to the performance and / or accuracy of the parameter allocation model may be sent from the first RAN node 504 and / or the second RAN node 506. Such feedback may comprise one or more measurement reports after taking an action determined by the parameter allocation model (e.g., applying a new setting of at least one radio parameter) and may be indicative of the performance of the trained parameter allocation model. Feedback provided to the 0AM 508 may be used for further model training and / or refinement in order to improve the performance of the parameter allocation model.

[0188] One or more of the steps and / or procedures described above with reference to FIG. 5 may be performed a plurality of times. For example, settings of at least one radio parameter may be determined and allocated several times during connection of a UE to the network. Furthermore, a parameter allocation model may undergo training and / or refinement several times based on newly collected data.

[0189] FIG. 6 is a signal flow diagram of further steps of an example method of allocating radio parameters for a user equipment. The method of FIG. 6 is performed by the same components and functions as were described above with referent to FIG. 5. The method shown in FIG. 6 may be performed after or as an optional part of the method described above with reference to FIG. 5.

[0190] The method of FIG. 6 begins with radio parameter configuration at step 532. Step 532 of the method of FIG. 6 corresponds with step 532 as described above with reference to FIG. 5 and no further detailed explanation of this step is provided again with reference to FIG. 6.

[0191] At step 602 a cell handover of the UE 502 from the first RAN node 504 to the second RAN node 506 is performed. The cell handover comprises reassigning a serving cell of the UE 502 from a cell operated by the first RAN node 504 to a cell operated by the second RAN node 506. The cell handover may occur because the UE 502 moves into a geographic region which is better served by the second RAN node 506 and / or because it is determined that improved performance may be achieved by performing the handover procedure.

[0192] At step 604 a parameter transfer is performed to transfer one or more parameters from the first RAN node 504 to the second RAN node 506. The parameter transfer may comprise sending one or more parameters such as a recently determined setting of at least one radio parameter for the UE 502 and / or one or more parameters of a trained parameter allocation model which was used to determine a setting of the at least one radio parameter. Transferring these one or more parameters to the second RAN node 506 may assist the second RAN node 506 in determining future settings of at least one radio parameter for the UE 502. For example, the parameters transferred in step 604 may allow the second RAN node 506 to benefit from training and / or configuration which has previously been performed for the UE 502.

[0193] At step 606 a further re-configuration of at least one radio parameter is performed for the UE 502 whilst the UE 502 is attached to the second RAN node 506. The radio parameter re-configuration of step 606 may include one or more steps corresponding to steps 524-532 of FIG. 5 but utilising the second RAN node 506 rather than the first RAN node 504. At least one aspect of the radio parameter re-configuration of step 606 may use the parameters provided in the parameter transfer of step 604.

[0194] In the examples presented above with reference to FIG. 5 and FIG. 6, the training of a parameter allocation model is performed at the 0AM 508 and model inference is performed at a RAN node (e.g.. the first RAN node 504). However, in other examples, one or more of these actions may be implemented by one or more other devices, network nodes and / or network functions. For example, model training may be implemented at one or more RAN nodes (e.g., the first RAN node 504 and / or second RAN node 506). Additionally or alternatively, model inference may be carried out at an 0AM 508. In some examples, the functionality of a RAN node (e.g., a gNB) may be split between a Centralized Unit (CU) or gNB-CU and a Distributed Unit (DU) or gNB-DU. In such examples, model inference may be implemented by a gNB-CU and model training may be implemented by an 0AM or a gNB-CU. In some examples, the model inference may be implemented by a node provided outside the RAN (for example, the model inference may be implemented in a Cloud network), and model training may be implemented by a RAN node such as a gNB. In some examples, both the model inference and the model training may be implemented outside the RAN (for example, in a Cloud network), and a scheduled radio parameter output by the model may be provided to a gNB.

[0195] Various methods have been described herein in which some of the method steps may be implemented on any suitable electronic device (such as a computing device) and / or combination of electronic devices (e.g. computing devices). FIG. 7 is a schematic illustration of an example electronic device 702 which may be used to implement all or part of any method described herein. For example, an electronic device 702 of the type shown in FIG. 7 and as described below may be used to implement all or part of a UE, a RAN node (e.g., a base station) and / or an 0AM as described herein.

[0196] The electronic device 702 may include at least one processing unit 704, memory 708 and an input / output interface 706 (I / O). The processing unit 704 may include any suitable processor and / or combination of processors. For example, the processing unit 704 may include one or more of a Central Processing Unit (CPU) and a Graphical Processing Unit (GPU). The memory 708 may include volatile memory and / or non-volatile / persistent memory. The memory 708 may, for example, be used to store data such as an operating system, instructions to be executed by the processing unit (e.g. in the form of software to be executed by the processing unit), configuration information related to the electronic device 702, session information and / or configuration or registration information associated with any other device, node or module in the network. In some examples, the memory 708 may be used to store instructions for executing any of the methods disclosed herein.

[0197] At least the processing unit 704 is connected to the input / output interface 706. The input / output interface 706 may facilitate communication with one or more other devices. For example, the input / output interface 706 may be operable to transmit and / or receive communications to / from other devices in a network. The input / output interface 706 may, for example, comprise one or more antennas to facilitate wireless communication with one or more other devices.

[0198] Optionally, the electronic device 702 may further include a display (not shown). The display may comprise any suitable electronic display such as a touch sensitive display. The display may be connected to at least to the processing unit 704. The processing unit 704 may generate display signals which are sent to the display in order to cause the display information.

[0199] Features, integers, characteristics or groups described in conjunction with a particular aspect, embodiment or example of the invention are to be understood to be applicable to any other aspect, embodiment or example described herein unless incompatible therewith. All of the features disclosed in this specification (including any accompanying claims, abstract and drawings), and / or all of the steps of any method or process so disclosed, may be combined in any combination, except combinations where at least some of such features and / or steps are mutually exclusive. The invention is not restricted to the details of any foregoing embodiments. The invention extends to any novel one, or any novel combination, of the features disclosed in this specification (including any accompanying claims, abstract and drawings), or to any novel one, or any novel combination, of the steps of any method or process so disclosed. In particular, any dependent claims may be combined with any of the independent claims and any of the other dependent claims.

[0200] Each feature disclosed in this specification (including any accompanying claims, abstract and drawings), may be replaced by alternative features serving the same, equivalent or similar purpose, unless expressly stated otherwise. Thus, unless expressly stated otherwise, each feature disclosed is one example only of a generic series of equivalent or similar features. The invention is not restricted to the details of any foregoing embodiments. The invention extends to any novel one, or any novel combination, of the features disclosed in this specification (including any accompanying claims, abstract and drawings), or to any novel one, or any novel combination, of the steps of any method or process so disclosed. The claims should not be construed to cover merely the foregoing embodiments, but also any embodiments which fall within the scope of the claims.

Claims

1. A method of allocating radio parameters for a first user equipment, UE, in a wireless communication network, the method comprising:receiving radio parameter data indicative of a first setting of at least one radio parameter being used for the first UE whilst the first UE provides a service using the wireless communication network;receiving quality of experience, QoE, data from the first UE, wherein the QoE data is indicative of a quality of a user experience provided by the first UE whilst the first UE provides the service using the wireless communication network and using the first setting for the at least one radio parameter;providing the received radio parameter data and the received QoE data to a parameter allocation model, wherein the parameter allocation model is configured to determine, in dependence on the radio parameter data and the QoE data, a second setting of the at least one radio parameter for use for the first UE in the wireless communication network and for the first UE to receive at least a prescribed QoE whilst using the second setting of the at least one radio parameter; andsending the second setting of the at least one radio parameter, determined by the parameter allocation model, to the first UE.

2. The method of claim 1, wherein the at least one radio parameter comprises a plurality of radio parameters, and wherein the first setting and the second setting comprise settings of each of the plurality of radio parameters.

3. The method of claim 1 or 2, wherein the at least one radio parameter comprises at least one radio parameter for which a setting of the radio parameter is allocated for use for the first UE for a plurality of instances of data transmission and / or reception.

4. The method of any one of claims 1 to 3, wherein the at least one radio parameter comprises at least one radio parameter for which a setting of the radio parameter is allocated for use for the first UE for a plurality of transmission time intervals.

5. The method of any one of claims 1 to 4, wherein sending the second setting of the at least one radio paramater to the first UE comprises sending the second setting to the first UE as part of a Radio Resource Control, RRC, message.

6. The method of any one of claims 1 to 5, wherein the at least one radio parameter comprises at least one of a parameter related to the configuration of a Hybrid Automatic Repeat Requestprocedure, a parameter related to the configuration of a discontinuous reception procedure, a parameter related to the configuration of a discontinuous transmission procedure, a parameter related to the configuration of a procedure carried out using a Medium Access Control protocol, a parameter related to the configuration of a procedure carried out using a Radio Link Control protocol, and / or a parameter related to the configuration of a procedure carried out using a Packet Data Convergence Protocol.

7. The method of any one of claims 1 to 6, wherein the parameter allocation model is configured through training based on a plurality of training data records, wherein each training data record comprises: a historical setting of the at least one radio parameter and measured QoE data, wherein the measured QoE data is indicative of a quality of a user experience provided by a UE whilst the UE provided a service using the wireless communication network and using the historical setting for the at least one radio parameter.

8. The method of any one of claims 1 to 7, wherein the parameter allocation model comprises a predictive model, configured through training to determine a predicted QoE which is expected to result from an input setting of the at least one radio parameter.

9. The method of claim 8, wherein the parameter allocation model is further configured to determine the second setting of the at least one radio parameter as a setting of the at least one radio parameter which the predictive model expects to result in at least a prescribed QoE whilst providing a service at the first UE and using the second setting of the at least one radio parameter.

10. The method of any one of claims 1 to 9, wherein the parameter allocation model is configured through training comprising a reinforcement learning process and wherein determining the second setting of the at least one radio parameter comprises an iteration of the reinforcement learning process.

11. The method of claim 10, wherein the reinforcement learning process is configured to maximise a reward comprising at least a measure of a QoE provided by the first UE whilst the first UE provides a service using the wireless communication network.

12. The method of any one of claims 1 to 11, wherein the method further comprises:receiving second radio parameter data indicative of a plurality of settings of the at least one radio parameter being used for a plurality of second UEs whilst the plurality of second UEs provide a service using the wireless communication network;receiving second quality of experience, QoE, data from the second UEs, wherein the second QoE data is indicative of a quality of a user experience provided by the second UEs whilst the second UEs provide the service using the wireless communication network and using the plurality of settings for the at least one radio parameter; andproviding the second received radio parameter data and the second received QoE data to the parameter allocation model, wherein the parameter allocation model is configured to determine the second setting of the at least one radio parameter for use by the first UE in further dependence on the second received radio parameter data and the second received QoE data.

13. The method of claim 12, wherein the parameter allocation model is configured to determine the second setting of the at least one radio parameter for use for the first UE and the second plurality of UEs to receive at least a prescribed QoE whilst the second setting of the at least one radio parameter is used for the first UE.

14. The method of claim 12 or 13, wherein the plurality of second UEs include at least some UEs attached to a plurality of Radio Access Network nodes of the wireless communication network.

15. The method of any one of claims 1 to 14, further comprising receiving radio coverage data from the first UE, wherein the radio coverage data is indicative of radio coverage provided to the first UE whilst the first UE provides the service using the wireless communication network and using the first setting for the at least one radio parameter,wherein the parameter allocation model is configured to determine the second setting of the at least one radio parameter for use by the first UE in further dependence on the received radio coverage data.

16. The method of any one of claims 1 to 15, further comprising:performing a cell handover for the first UE, wherein the cell handover comprises handover of the first UE from a first cell operated by a first Radio Access Network, RAN, node to a second cell operated by a second RAN node; andsending the determined second setting of the at least one radio parameter to the second RAN node for use by the second RAN node.

17. A method of training a radio parameter allocation model for determining a setting of at least one radio parameter for use for a UE in a wireless communication network, the method comprising:receiving a plurality of training data records, wherein each training data record comprises:an historical setting of the at least one radio parameter used for a UE in the wireless communication network; andQuality of Experience, QoE, measurement data indicative of a quality of a user experience provided by the UE whilst the UE provided a service using the wireless communication network and using the historical setting of the at least one radio parameter; andtraining the parameter allocation model, in dependence on the received training data records, for determining a setting of at least one radio parameter for use for a UE in the wireless communication network and for the UE to receive at least a prescribed QoE whilst using the determined setting.

18. A computer-readable medium comprising instructions which, when executed by a computing apparatus, cause the computing apparatus to carry out the method of any one of claims 1 to 17.

19. A computing system for use in a wireless communication network, the system configured to:receive radio parameter data indicative of a first setting of at least one radio parameter being used for a first UE whilst the first UE provides a service using the wireless communication network;receive quality of experience, QoE, data from the first UE, wherein the QoE data is indicative of a quality of a user experience provided by the first UE whilst the first UE provides the service using the wireless communication network and using the first setting for the at least one radio parameter;provide the received radio parameter data and the received QoE data to a parameter allocation model, wherein the parameter allocation model is configured to determine, in dependence on the radio parameter data and the QoE data, a second setting of the at least one radio parameter for use for the first UE in the wireless communication network and for the first UE to receive at least a prescribed QoE whilst using the second setting of the at least one radio parameter; andsend the second setting of the at least one radio parameter, determined by the parameter allocation model, to the first UE.

20. A computing system for use in a wireless communication network, the system configured to:receive a plurality of training data records, wherein each training data record comprises:an historical setting of the at least one radio parameter used for a UE in the wireless communication network; andQuality of Experience, QoE, measurement data indicative of a quality of a user experience provided by the UE whilst the UE provided a service using the wireless communication network and using the historical setting of the at least one radio parameter; andtrain a parameter allocation model, in dependence on the received training data records, for determining a setting of at least one radio parameter for use for a UE in the wireless communication network and for the UE to receive at least a prescribed QoE whilst using the determined setting.

Images