Improving network decision in a cellular telecommunication system

Abstract

A user equipment comprising: means for sending to a network node an indicator configured to map a user equipment state to a network traffic prediction for the user equipment. A network apparatus comprising: means for receiving from a user equipment an indicator configured to map a user equipment state to a network traffic prediction for the user equipment.

Description

[0001] IMPROVING NETWORK DECISION IN A CELLULAR TELECOMMUNICATION

[0002] SYSTEM

[0003] TECHNOLOGICAL FIELD

[0004] Examples of the disclosure relate to improving network decisions in a cellular telecommunication system.

[0005] BACKGROUND

[0006] In cellular telecommunications systems, the network generally controls operation such as allocation of resources. It is desirable for the network to make ‘good’ decisions. This requires that the network has sufficient information in a form that allows the network to process that information efficiently.

[0007] BRIEF SUMMARY

[0008] According to various, but not necessarily all, examples there is provided examples as claimed in the appended claims.

[0009] While the above examples of the disclosure and optional features are described separately, it is to be understood that their provision in all possible combinations and permutations is contained within the disclosure. It is to be understood that various examples of the disclosure can comprise any or all of the features described in respect of other examples of the disclosure, and vice versa. Also, it is to be appreciated that any one or more or all of the features, in any combination, may be implemented by / comprised in / performable by an apparatus, a method, and / or computer program instructions as desired, and as appropriate.

[0010] BRIEF DESCRIPTION

[0011] Some examples will now be described with reference to the accompanying drawings in which:

[0012] FIG. 1 shows an example of a network;

[0013] FIG. 2 shows an example of the subject matter described herein;

[0014] FIG. 3 shows an example of the subject matter described herein;

[0015] FIG. 4A shows an example of the subject matter described herein;

[0016] FIG. 4B shows an example of the subject matter described herein; FIG. 5 shows an example of the subject matter described herein;

[0017] FIG. 6A shows an example of the subject matter described herein;

[0018] FIG. 6B shows an example of the subject matter described herein;

[0019] FIG. 7A shows an example of the subject matter described herein;

[0020] FIG. 7B shows an example of the subject matter described herein;

[0021] FIG. 8 shows an example of the subject matter described herein;

[0022] FIG. 9 shows an example of the subject matter described herein;

[0023] The figures are not necessarily to scale. Certain features and views of the figures can be shown schematically or exaggerated in scale in the interest of clarity and conciseness. For example, the dimensions of some elements in the figures can be exaggerated relative to other elements to aid explication. Similar reference numerals are used in the figures to designate similar features. For clarity, all reference numerals are not necessarily displayed in all figures.

[0024] In the following description a class (or set) can be referenced using a reference number without a subscript index (e.g. 10) and a specific instance of the class (member of the set) can be referenced using the reference number with a numerical type subscript index (e.g. 10_1 ) and a non-specific instance of the class (member of the set) can be referenced using the reference number with a variable type subscript index (e.g. 10J).

[0025] DETAILED DESCRIPTION

[0026] FIG 1 illustrates an example of a network 100 comprising a plurality of network nodes including terminal nodes 110, access nodes 120 and one or more core nodes 129.

[0027] The terminal nodes 110 and access nodes 120 communicate with each other. The one or more core nodes 129 communicate with the access nodes 120.

[0028] The network 100 is in this example a radio telecommunications network, in which at least some of the terminal nodes 110 and access nodes 120 communicate with each other using transmission / reception of radio waves.

[0029] The one or more core nodes 129 may, in some examples, communicate with each other. The one or more access nodes 120 may, in some examples, communicate with each other. The network 100 may be a cellular network comprising a plurality of cells 122 each served by an access node 120. In this example, the interface between the terminal nodes 110 and an access node 120 defining a cell 122 is a wireless interface 124.

[0030] The access node 120 is a cellular radio transceiver. The terminal nodes 110 are cellular radio transceivers.

[0031] In the example illustrated the cellular network 100 is a third generation Partnership Project (3GPP) network in which the terminal nodes 110 are user equipment (UE) and the access nodes 120 are base stations.

[0032] In the particular example illustrated the network 100 is an Evolved Universal Terrestrial Radio Access network (E-UTRAN). The E-UTRAN consists of E-UTRAN NodeBs (eNBs) 120, providing the E-UTRA user plane and control plane (RRC) protocol terminations towards the UE 110. The eNBs 120 are interconnected with each other by means of an X2 interface 126. The eNBs are also connected by means of the S1 interface 128 to the Mobility Management Entity (MME) 129.

[0033] In another example the network 100 is a Next Generation (or New Radio, NR) Radio Access network (NG-RAN). The NG-RAN consists of gNodeBs (gNBs) 120, providing the user plane and control plane (RRC) protocol terminations towards the UE 110. The gNBs 120 are interconnected with each other by means of an X2 / Xn interface 126. The gNBs are also connected by means of the N2 interface 128 to the Access and Mobility management Function (AMF).

[0034] A user equipment comprises a mobile equipment. Where reference is made to user equipment that reference includes and encompasses, wherever possible, a reference to mobile equipment.

[0035] In the following description reference will be made to a network node. The network node can be any suitable network node. For convenience the network node is labelled with reference numeral 120. In some examples, the network node is an access node 120, however, in other examples the network node is another network node, for example, a terminal node 1 10 or a core network entity.

[0036] FIG. 2 illustrates an example of a system comprising one or more user equipment 110 and one or more network nodes 120. The system can be part of a network 100.

[0037] The user equipment 110 comprises means for sending to the network node 120 an indicator 20 configured to map a user equipment state 10 to a network traffic prediction 40 for the user equipment 110.

[0038] The network node 120 comprises means for receiving from the user equipment 110 the indicator 20 configured to map the user equipment state 10 to the network traffic prediction 40 for the user equipment 110.

[0039] In the example illustrated, the user equipment state 10, S(UE) is mapped 30 by the user equipment 110 to an indicator 20, [I]. The indicator 20, [I], is transferred from the user equipment 110 to the network node 120 (NN). The received indicator 20, [I], is mapped 50 by the network node 120 to a network traffic prediction 40 for the user equipment 1 10, NTP(UE). Therefore, in the example illustrated, the user equipment state 10, S(UE) is mapped 30 to a network traffic prediction 40 for the user equipment 1 10, NTP(UE) via the transfer of the indicator 20.

[0040] The transferred indicator 20, [I], allows the network node 120 to determine a network traffic prediction 40 for the user equipment 1 10, NTP(UE). This can be used by the network to anticipate, control, or configure UE and / or network resources (e.g. grantsZallocation(s)) for network traffic 60. In some examples, the network traffic 60 is a sequence of traffic for the UE. In some examples, the network traffic 60 is or comprises downlink data traffic from the network node 120 to the user equipment 110. In some examples, the network traffic 60 is or comprises uplink data traffic from the user equipment 110 to the network node 120 and downlink data traffic from the network node 120 to the user equipment 110.

[0041] In the example illustrated, but not necessarily all examples, the network node 120 is configured to use 80 the determined network traffic prediction 40 for the user equipment 110, NTP(UE) to enable the network to make better radio resource management decisions in light of the user equipment 110 assisted prediction of network traffic arrivals for the user equipment 110. The network node 120 sends 80 a downlink resource configuration message 82 to the user equipment 110, for example, allocating resources for the network traffic (e.g. proactive grants) and proactively keeping the user equipment 110 awake.

[0042] In the example illustrated, but not necessarily all examples, the network node 120 is configured to use 80 the determined network traffic prediction 40 for the user equipment 110, NTP(UE) to configure data transfer between the network node 120 and the user equipment 110. The network node 120 sends a downlink resource configuration message 82 to the user equipment 110 allocating resources for the network (downlink) traffic.

[0043] In at least some examples, the network traffic prediction 40 for the user equipment 110 is valid for an observation period. The network traffic prediction 40 for the user equipment 110 predicts network traffic 60 over the observation period but not beyond the observation period. The network traffic 60 predicted can have a fixed or a time variable profile over the observation period. The network traffic 60 predicted can have one or more bursts of network traffic over the observation period. The network traffic 60 predicted can be substantially continuous over the observation period.

[0044] The user equipment state 10 is a precursor to the data transfer 60. The user equipment state 10 causes the indicator 20 to be transferred to the network node 120, where it is used to is used to obtain a network traffic prediction 40 for the user equipment 110.

[0045] The user equipment state 10 is a state associated with subsequent network traffic 60. This association may be learned as described later.

[0046] In some examples, the user equipment state 10 is a state of the user equipment 110 expected to cause data transfer within an observation period. Additionally or alternatively, in at least some examples, the user equipment state 10 is a software state associated with a state of an operating system (OS) of the user equipment 110 or associated with a state of an application running on the user equipment 110 or is a state of an artificial intelligence / machine learning (AIMNL) traffic prediction system

[0047] In some examples there is a AIML entity within the user equipment, configured to observe traffic patterns, and possibly other software state inputs, and can use these to correlate the software state inputs (if any) and prior / current traffic with the upcoming traffic pattern.

[0048] In some examples, the user equipment state 10 is a hash of application parameters for an application capable of transferring data between the user equipment 110 and the network node 120 that is running on the user equipment 110.

[0049] In some examples, the user equipment state 10 is a state triggered by a change in state of the operating system of the user equipment 110 or a change in state of an application running on the user equipment 110. Additionally or alternatively, in at least some examples, the user equipment state 10 is triggered by an action of a user of the user equipment 110.

[0050] In some examples, the user equipment 110 is configured to detect that a trigger condition for sending an indicator 20 is met; and, in response to detecting that the trigger condition for sending the indicator 20 has been met, causing sending to the network node 120 the indicator 20 configured to map the user equipment state 10 to the network traffic prediction 40 for the user equipment 110.

[0051] In some examples, the user equipment state 10 is a necessary and sufficient condition for the trigger condition to be met. In some examples, the user equipment state 10 is a necessary but not a sufficient condition and an additional condition is required to be met or additional conditions are required to be met before the trigger condition is met. For example, a sequence of user equipment states 10 may be required. For example, a user action such as a user input may be required.

[0052] In some examples, the likelihood of sending the indicator 20 configured to map a user equipment state 10 to a network traffic prediction 40 for the user equipment 110 is dependent upon the additional condition(s). In some examples, the additional condition(s) is related to an event expected to cause data transfer (UL / DL) within an observation period.

[0053] In some examples, the additional condition is an application layer condition.

[0054] In some examples, the additional condition is a user action condition e.g. gaze / click / tap, option choice, app launch, speech).

[0055] In some examples, the additional condition is alternatively a condition for a lower layer than the application layer.

[0056] In some examples, the additional condition is a background synchronization or is based on current or predicted user behaviour.

[0057] In some examples, the additional condition is expected maintenance / consistency of the software state 10 over an observation period.

[0058] The user equipment 110 distills information about the context of the user equipment 110 and in particular the user equipment state 10 and provides that distilled information via indicator 20. This information is not otherwise available to the network node 120 at that time. The network node 120 can then use that indicator 20 to quickly and timely determine a traffic pattern prediction 40 that can be used for better decision making, for example better resource allocation based on timely, accurate, forward looking predictions as regards an upcoming pattern of traffic. The resource allocation can for example reduce latency on uplink and / or downlink (for example, provide just in time uplink grant). The resources allocation can for example modify a user equipment (UE) discontinuous reception (DRX) active time for the user equipment 110.

[0059] In some but not necessarily all examples the user equipment 110 comprises means for receiving a configuration message 70 from the network node 120 that configures at the user equipment 110 sending of the indicator 20.

[0060] In some examples, the configuration message 70 is a radio resource control (RRC) message. In some examples, the configuration message 70 is a medium access control (MAC) message. In some examples, the network node 120 determines whether the configuration message 70 is a radio resource control (RRC) message or a medium access control (MAC) message.

[0061] As illustrated in FIG 3, in some examples, the configuration message 70 comprises one or more information elements 71.

[0062] In this example, the information element(s) 71 configure(s) the indicator 20; and / or configure(s) an observation period associated with the indicator 20 and / or configure(s) the user equipment state 10 associated with the trigger condition and / or configure(s) at least one additional (trigger) condition for sending the indicator 20 and / or configure(s) a likelihood of sending the indicator 20 configured to map a user equipment state 10 to a network traffic prediction 40 for the user equipment 110.

[0063] In at least some example the likelihood of sending the indicator is dependent upon a likelihood threshold or confidence level for sending the indicator 20.

[0064] In some examples, the configuration message 70 configures the user equipment 110 to: stop sending any indicator 20 or stop sending a specified indicator 20 or to change an observation period associated with an indicator 20 or to change a user equipment state 10 associated with an indicator 20 or to change a trigger condition causing sending of an indicator 20 or prevent a trigger condition causing sending of an indicator 20.

[0065] FIG 4A illustrates an example of a learning phase (mode) where a user equipment state 10 is associated with subsequent network traffic 60. The indicator 20 is associated, at the user equipment 110, with the user equipment state 10 and is associated, via learning 200 at the network node 120, with a network traffic prediction 40 for the user equipment 110.

[0066] The user equipment 110, creates a mapping 30 that associates the indicator 20 and the user equipment state 10.

[0067] The network node 120 is informed of the indicator 20 and creates a mapping 50 that associates the indicator 20 with a network traffic prediction 40 for the user equipment 110. The network traffic prediction 40 is based on the network traffic 60 associated with the user equipment state 10. In at least some examples, it is the network traffic 60 that occurs within an observation period following, possibly immediately following, the user equipment state 10.

[0068] In some examples, the indicator 20 is transferred from the user equipment 110 to the network node 120 before the network traffic 60 that occurs within an observation period following the user equipment state 10. In some examples, the indicator 20 is transferred from the user equipment 110 to the network node 120 after or during the network traffic 60 that occurs within an observation period following the user equipment state 10.

[0069] Where the indicator 20 is transferred from the user equipment 110 to the network node 120 before the network traffic 60 that occurs within an observation period following the user equipment state 10, the indicator 20 can be transferred from the user equipment 110 to the network node 120 in response to the user equipment state 10.

[0070] In at least some examples, when the indicator 20 is transferred from the user equipment 110 to the network node 120, the indicator 20 implicitly or explicitly indicates a start of the observation period.

[0071] The network traffic 60 for the user equipment can be measured by the network node 120 and used to determine a network traffic prediction 40.

[0072] The network traffic 60 for the user equipment 110 can be measured by the user equipment 110 and the network node 120 and used by the network node 120 to determine a network traffic prediction 40 which is used by the network node 120.

[0073] FIG 4B illustrates an example of an inference phase (mode) where a user equipment state 10 is used to predict subsequent network traffic 60. The user equipment state 10 (possibly in combination with other trigger conditions) causes the indicator 20 associated with the user equipment state 10 by the mapping 30 to be sent to the network node 120. The received indicator 20 is used by the network node 120 to obtain the network traffic prediction 40 mapped by the mapping 50 of the received indicator 20.

[0074] Thus as illustrated in FIG 4A, the user equipment 110 is configured to send to the network node 120 the indicator 20 before a mapping 50, enabling the indicator 20 to map the user equipment state 10 to the network traffic prediction 40 for the user equipment 110, has been established to enable establishment of the mapping 50. As illustrated in FIG 4B, the user equipment 110 is configured to send to the network node 120 the indicator 20 configured to map the user equipment state 10 to the network traffic prediction 40 for the user equipment 110 after the mapping has been established.

[0075] Thus as illustrated in FIG 4A, the network node 120 is configured to receive from the user equipment 110 before a mapping 50, enabling the indicator 20 to map the user equipment state 10 to the network traffic prediction 40 for the user equipment 110, has been established to enable establishment of the mapping 50. As illustrated in FIG 4B, the network node 120 is configured to receive from the user equipment 110 the indicator 20 configured to map the user equipment state 10 to the network traffic prediction 40 for the user equipment 110 after the mapping 50 has been established.

[0076] In some examples, the mapping 50 maps directly from the indicator 20 to the network traffic prediction 40 for the user equipment 110, for example, an input to the mapping 50 of the indicator 20 produces as an output a network traffic prediction 40 that is associated by the mapping with the input indicator 20.

[0077] In some examples, the mapping 50 maps indirectly from the indicator 20 to the network traffic prediction 40 for the user equipment 110, for example, an input to the mapping 50 of the indicator 20 produces an output parameter that is associated by the mapping with the input indicator 20 and the output parameter is associated with network traffic prediction 40.

[0078] Thus the mapping 50, for example, is received as an input the indicator 20 and produces as an output which may or may not be described using the words “network traffic prediction” but will have the qualities of predicting or enabling the prediction of network traffic 60.

[0079] The above-described learning can be achieved using training. The above-described prediction can be achieved using inference.

[0080] In at least some examples, the user equipment 110 is configured to enter a learning mode configured by the network node 120 to create the mapping 50 that maps, directly or indirectly, from the indicator 20 to the network traffic prediction 40 for the user equipment 110.

[0081] In some examples, the learning phase (mode) comprises training a machine learning model at the network node 120 using training data, at least some of which is provided by the user equipment 110. The training is realized on multiple samples (training dataset).

[0082] The trained machine learning model provides the mapping 50. A machine learning model can, for example, be trained using gradient descent of a loss function. The loss function represents a loss between the output of the machine learning model and the target output.

[0083] The machine learning model can, for example, be an artificial neural network (ANN) comprising layers of multiple artificial neurons. Each artificial neuron is connected to receive an input from multiple artificial neurons in the previous layer via weights specific to the input. The weights provide different weights to the inputs received from the artificial neurons in the previous layer. The weights are estimated for specific input and corresponding output. The values for the weights can be trained using gradient descent, for example, using back-propagation.

[0084] The training of the machine learning model requires large amounts of training data. This data is input to the machine learning model to train the machine learning model. For example, it is used to generate the loss function values used to train the machine learning model via gradient descent. In this example, the machine learning model is trained to receive as an input an indicator 20 and to produce as an output a network traffic prediction 40.

[0085] During training the machine learning model uses labelled training data. It is normal for this to be received and processed in batches.

[0086] A label indicates the measured network traffic 60 associated with the training data (the indicator 20 associated with that measured network traffic 60). The machine learning model learns to produce an output for an input indicator 20 similar to the associated measured network traffic 60 this similarity provides a predictive function. The machine learning model consequently learns to produce as an output for an input indicator 20, a network traffic prediction 40 that predicts expected network traffic 60.

[0087] In this example, the machine learning model is trained to receive as an input an indicator 20 and to produce as an output a network traffic prediction 40 for the user equipment.

[0088] In some examples, a machine learning model is trained for each user equipment separately. A user equipment identifier is used to discriminate between user equipment and is used to select the appropriate machine learning model during training and inference. The user equipment identifier can be sent from the user equipment 110 to the network node 120.

[0089] In other examples, a machine learning model is trained for groups of user equipment separately. A group identifier is used to discriminate between groups of user equipment and is used to select the appropriate machine learning model during training and inference. The group identifier can, for example, indicate one or more of: a manufacturer of the user equipment 110, an equipment-model of the user equipment 110, an operating system of the user equipment 110, an in-use application at the user equipment 110, a service level associated with the user equipment 110 etc. The group identifier can be sent from the user equipment 110 to the network node 120. In other examples, a machine learning model is trained for all user equipments. The above described user equipment identifier and / or group identifiers and / or serving cell identifier can be used as training data and input data during inference. The machine learning model learns which inputs are of importance to convert an input identifier 20 to a robust network traffic prediction 40.

[0090] An advantage of a machine learning model is that the machine learning model can determine which of various inputs are important and how they are of importance to convert an input identifier 20 to a robust network traffic prediction 40.

[0091] The network node 120 can then use the determined network traffic prediction 40 for better decision making, for example better resource allocation based on timely, accurate, forward looking predictions as regards an upcoming pattern of traffic. The resource allocation can for example reduce latency on uplink and / or downlink (for example, provide just in time uplink grant). The resources allocation can for example modify an active time for the user equipment 110.

[0092] The improvement achieved by the decision-making based on the determined network traffic prediction 40 from a received indicator 20, can be assessed at the network node 120 and / or the user equipment 110. For example, the user equipment can send to the network node 120 information on delays at the user equipment 110 in association with the indicator 20. The network node 120 can then use the assessment as a feedback signal for improving resource allocation based on the network traffic prediction 40 determined from the received indicator 20.

[0093] In some examples, during the training phase, the network node 120 provisions additional proactive uplink grants, and / or scheduling requests (SRs) to enable the network node 120 to more precisely learn the traffic pattern associated with an indicator 20.

[0094] The learning phase (node) can continue in parallel with the inference phase (mode). For both training and inference the user equipment sends an indicator 20. In the inference mode, the indicator 20 is converted to a network traffic predictor to predict future network traffic. In the learning mode, the indicator 20 is used to label past or future network traffic. During the inference mode, the indicator 20 can be mapped to a network traffic predictor to predict future network traffic (as described) and can additionally be used to label past or future network traffic during an on-going learning phase. It is therefore possible for the network to make changes and observe the impact of those changes on the predicted network traffic and the actual network traffic. This allows the network node to obtain convergence (minimize the prediction error) between the predicted network traffic and the actual network traffic. Thus in some examples, the network node and user equipment are configured to perform extended learning, where the learning extends into the inference phase and the indicator received by the network node for inference is also used for learning.

[0095] The predicted network traffic associated with a received indicator 20 can evolve over time as the mapping 50 is updated by learning. The network traffic pattern associated with the received indicator 20 is not necessarily constant each and every time that indicator 20 is received because it can also react and evolve to the network traffic actually observed subsequent to the (initial / earlier) reception of the indicator 20.

[0096] FIG 5 illustrates an example of a learning phase (mode) where first user equipment state 10_1 is associated with subsequent first network traffic 60_1 and a second user equipment state 10_2 is associated with subsequent second network traffic 60_2.

[0097] The first indicator 20_1 is associated, at the user equipment 110, with the first user equipment state 10_1 and is associated, via learning 200_1 at the network node 120, with a first network traffic prediction 40_1 for the user equipment 110.

[0098] The user equipment 110, creates a first mapping 30_1 that associates the first indicator 20_1 and the first user equipment state 10_1 .

[0099] The network node 120 is informed of the first indicator 20_1 and creates a first mapping 50_1 that associates the first indicator 20_1 with the first network traffic prediction 40_1 for the user equipment 110.

[0100] The first network traffic prediction 40_1 is based on the first network traffic 60_1 associated with the first user equipment state 10_1 . In at least some examples, it is the network traffic that occurs within an observation period immediately following the first user equipment state 10_1 .

[0101] The second indicator 20_2 is associated, at the user equipment 110, with the second user equipment state 10_2 and is associated, via learning 200_2 at the network node 120, with a second network traffic prediction 40_2 for the user equipment 110.

[0102] The user equipment 110, creates a second mapping 30_2 that associates the second indicator 20_2 and the second user equipment state 10_2.

[0103] The network node 120 is informed of the second indicator 20_2 and creates a second mapping 50_2 that associates the second indicator 20_2 with the second network traffic prediction 40_2 for the user equipment 110.

[0104] The second network traffic prediction 40_2 is based on the second network traffic 60_2 associated with the second user equipment state 10_2. In at least some examples, it is the network traffic that occurs within an observation period immediately following the second user equipment state 10_2.

[0105] As described previously with reference to FIG 3, an indicator 20_1 , 20_2 can be transferred from the user equipment 110 to the network node 120 before / after / during the network traffic 60 that occurs within an observation period following the user equipment state 10, and can be transferred in response to the user equipment state 10 (with or without additional trigger conditions). An indicator 20_1 , 20_2 can implicitly or explicitly indicate a start of a respective observation period. The network traffic 60 for the user equipment 110 can be measured by the network node 120 and / or the user equipment 110.

[0106] Referring additionally to FIG 6A, the first user equipment state 10_1 is a state of the user equipment 110 that causes first data transfer 60_1 within a first observation period 90_1 . The first user equipment state 10_1 is a state of the user equipment 110 expected to cause similar first data transfer 60_1 within a similar observation period 90_1 . The expected similar first data transfer 60_1 within a similar first observation period 90_1 is identified by the first network traffic prediction 40_1 . The first learning 200_1 at the network node 120 associates the first indicator 20_1 with a first network traffic prediction 40_1 for the user equipment 110. The first network traffic prediction 40_1 for the user equipment 110 predicts network traffic 60_1 over the first observation period 90_1 but not beyond the first observation period 90_1 .

[0107] The first network traffic 60_1 used for learning and prediction can have a fixed or a time variable profile over the first observation period 90_1 as illustrated in FIG 6A. The first network traffic 60_1 used for learning and prediction can have one or more bursts 62_1 of network traffic 60_1 over the first observation period 90_1 . The first network traffic 60_1 used for learning and prediction can be substantially continuous (or not) over the first observation period 90_1 .

[0108] Referring to FIG 6B, the second user equipment state 10_2 is a state of the user equipment 110 that causes second data transfer 60_2 within a second observation period 90_2. The second user equipment state 10_2 is a state of the user equipment 110 expected to cause similar second data transfer 60_2 within a similar second observation period 90_2. The expected similar second data transfer 60_2 within a similar second observation period 90_2 is the second network traffic prediction 40_2. The second learning 200_2 at the network node 120 associates the second indicator 20_2 with a second network traffic prediction 40_2 for the user equipment 110. The second network traffic prediction 40_2 for the user equipment 110 predicts network traffic 60_2 over the second observation period 90_2 but not beyond the second observation period 90_2.

[0109] The second network traffic 60_2 used for learning and prediction can have a fixed or a time variable profile over the second observation period 90_2 as illustrated in FIG 6B. The second network traffic 60_2 used for learning and prediction can have one or more bursts 62_2 of network traffic 60_2 over the second observation period 90_2. The second network traffic 60_2 used for learning and prediction can be substantially continuous (or not) over the second observation period 90_2.

[0110] FIGs 7A and 7B illustrate an example of an inference phase (mode) where a user equipment state 10 is used to predict subsequent network traffic 60. The user equipment state 10 (possibly in combination with other trigger conditions) causes the indicator 20 associated with the user equipment state 10 by the mapping 30 to be sent to the network node 120. The received indicator 20 is used by the network node 120 to obtain the network traffic prediction 40 mapped by the mapping 50 to the received indicator 20.

[0111] In FIG 7A, the first user equipment state 10_1 (possibly in combination with other first trigger conditions) causes the first indicator 20_1 associated with the first user equipment state 10_1 by the first mapping 30_1 to be sent to the network node 120. The received first indicator 20_1 is used by the network node 120 to obtain the first network traffic prediction 40_1 mapped by the first mapping 50_1 to the received first indicator 20_1 .

[0112] In FIG 7B, the second user equipment state 10_2 (possibly in combination with other second trigger conditions) causes the second indicator 20_2 associated with the second user equipment state 10_2 by the second mapping 30_2 to be sent to the network node 120. The received second indicator 20_2 is used by the network node 120 to obtain the second network traffic prediction 40_2 mapped by the second mapping 50_2 to the received second indicator 20_2.

[0113] Thus as illustrated in FIG 5, the user equipment 110 is configured to send to the network node 120 the first indicator 20_1 before a first mapping 50_1 , enabling the first indicator 20_1 to map the first user equipment state 10_1 to the first network traffic prediction 40_1 for the user equipment 110, has been established to enable establishment of the first mapping 50_1 . As illustrated in FIG 7A, the user equipment 110 is configured to send to the network node 120 the first indicator 20_1 configured to map the first user equipment state 10_1 to the first network traffic prediction 40_1 for the user equipment 110 after the first mapping 50_1 has been established.

[0114] Thus as illustrated in FIG 5, the network node 120 is configured to receive from the user equipment 110 before a first mapping 50_1 , enabling the first indicator 20_1 to map the first user equipment state 10_1 to the first network traffic prediction 40_1 for the user equipment 110, has been established to enable establishment of the first mapping 50_1. As illustrated in FIG 7A, the network node 120 is configured to receive from the user equipment 110 the first indicator 20_1 configured to map the first user equipment state 10_1 to the first network traffic prediction 40_1 for the user equipment 110 after the first mapping 50_1 has been established.

[0115] Thus as illustrated in FIG 5, the user equipment 110 is configured to send to the network node 120 the second indicator 20_2 before a second mapping 50_2, enabling the second indicator 20_2 to map the second user equipment state 10_2 to the second network traffic prediction 40_2 for the user equipment 110, has been established to enable establishment of the second mapping 50_2. As illustrated in FIG 7B, the user equipment 110 is configured to send to the network node 120 the second indicator 20_2 configured to map the second user equipment state 10_2 to the second network traffic prediction 40_2 for the user equipment 110 after the second mapping 50_2 has been established.

[0116] Thus as illustrated in FIG 5, the network node 120 is configured to receive from the user equipment 110 before a second mapping 50_2, enabling the second indicator 20_2 to map the second user equipment state 10_2 to the second network traffic prediction 40_2 for the user equipment 110, has been established to enable establishment of the second mapping 50_2. As illustrated in FIG 7B, the network node 120 is configured to receive from the user equipment 110 the second indicator 20_2 configured to map the second user equipment state 10_2 to the second network traffic prediction 40_2 for the user equipment 110 after the second mapping 50_2 has been established.

[0117] The first mapping 50_1 can map directly or indirectly to the first network traffic prediction 40_1 as previously described. The second mapping 50_2 can map directly or indirectly to the second network traffic prediction 40_2 as previously described.

[0118] A previously described, the first mapping 50_1 and / or the second mapping 50_1 can be produced using machine learning. In some examples, the learning phase (mode) illustrated in FIG 5 comprises training one or more machine learning models at the network node 120 using training data, at least some of which is provided by the user equipment 110. The trained machine learning model(s) provides the first mapping 50_1 and the second mapping 50_2. The machine learning model can, for example, be an artificial neural network (ANN). The machine learning model(s) is / are trained to receive as an input a first indicator 20_1 and to produce as an output a first network traffic prediction 40_1 and to receive as an input a second indicator 20_2 and to produce as an output a second network traffic prediction 40_2.

[0119] As previously described, there is flexibility to using machine learning. A machine learning model can be trained for each user equipment separately, for groups of user equipment separately, for all user equipment. User equipment identifiers and / or group identifiers can be used as additional input to the machine learning model..

[0120] The configuration message 72 sent to the user equipment 110 can, for example, configure the user equipment 110 to assist with training machine learning model(s) at the network node 120. The configuration message 72 can, for example, indicate that training data is required to train machine learning model(s) and / or what training data is required to train machine learning model(s).

[0121] In some examples, the configuration message 72 from the network node 120 configures at the user equipment 110 machine learning model training. The configuration message 72 can, for example, indicate what training data is required to train a machine learning model.

[0122] The configuration message(s) 72 sent from the network node 120 to the user equipment to configure the user equipment for the learning phase (mode) can be a radio resource control (RRC) message(s) and / or a medium access control (MAC) message(s) as previously described. In some examples, the configuration message 72 comprises one or more information elements 71 as previously described and / or configures the user equipment 110 as previously described.

[0123] The configuration in respect of the first mapping 50_1 between the first indicator 20_1 and the first network traffic prediction 40_1 can be independent of the configuration in respect of the second mapping 50_2 between the second indicator 20_1 and the second network traffic prediction 40_2. Although FIGs 5, 6A, 6B, 7A, 7B illustrate a first mapping 50_1 between the first indicator 20_1 and the first network traffic prediction 40_1 and a second mapping 50_2 between the second indicator 20_1 and the second network traffic prediction 40_2, the example can be extended to a mapping 50J between the indicator 20_i and the network traffic prediction 40J , where i is any number greater than 2.

[0124] It will therefore be appreciated that in at least some examples (FIG 7A), the user equipment 110 is configured to send to the network node 120 at a first time a first indicator 20_1 configured to map 50_1 a first user equipment state 10_1 to a first network traffic prediction 40_1 for the user equipment 110 UE and the user equipment 110 is configured to send to the network node 120 at a second time (different to the first time) a second indicator 20_2 (different to the first indicator 20_1) configured to map 50_2 a second user equipment state 10_2 (different to the first user equipment state 10_1 ) to a second network traffic prediction 40_2 for the user equipment 110.

[0125] In some examples, the configuration message 70, 72 sent to the user equipment to configure learning (training) and / or prediction (inference) comprises one or more information elements configured to configure a plurality of indicators 20 including the first indicator 20_1 and a second indicator 20_2.

[0126] In some examples, the configuration message 70, 72 sent to the user equipment to configure learning and / or inference comprises one or more information elements configured to: configure one or more observation periods associated with the plurality of indicators 20 including the first indicator 20_1 and the second indicator 20_2; and / or configure multiple user equipment states 10 including the first user equipment state 10_1 and / or the second user equipment state; configure trigger conditions for sending an indicator 20 configured to map a user equipment state 10 to a network traffic prediction 40 including a first trigger condition for sending the first indicator 20 and a second trigger condition for sending the second indicator 20. In some examples during prediction (inference) phase, the user equipment transmits multiple different indicators 20_1 , 20_2, with corresponding likelihoods. For example, the user equipment 110 sends the first indicator 20_1 with likelihood 70% if the user equipment 110 has 70% confidence that the first indicator 20_1 is the correct indicator. For example, the user equipment 110 sends the second indicator 20_2 with likelihood 30% if the user equipment 110 has 30% confidence that the second indicator 20_2 is the correct indicator.

[0127] In some examples, the first indicator 20_1 , but not the second indicator 20_2, is dependent upon an identifier of the user equipment 110 previously transferred to the network node 120 by the user equipment 110. For example, the second indicator 20_2, but not the first indicator 20_1 , is dependent upon a group identifier previously transferred to the network node 120 by the user equipment 110.

[0128] In the foregoing examples, a user equipment 110 comprises: means for sending to a network node an indicator 20 configured to map a user equipment state 10 to a network traffic prediction 40 for the user equipment 110.

[0129] The user equipment 110 comprises means for sending to a network node, during a learning phase (mode) an indicator 20 (a training indicator) configured to create a map from a user equipment state 10 to a network traffic prediction 40 for the user equipment 110. The user equipment 110 comprises means for sending to a network node, during an inference phase (mode) an indicator 20 (a predictive indicator) configured to map a user equipment state 10 to a network traffic prediction 40 for the user equipment 110. In some but not necessarily all examples, the indicator 20 comprises or conveys information that identifies it as a training indicator. In some additional or alternative examples, the indicator 20 comprises or conveys information that identifies it as a predictive indicator.

[0130] The network traffic prediction 40 relates to a prediction of a sequence of traffic for the UE (traffic arrivals at the network node 120). The network traffic prediction 40 relates to the traffic load. This is more specific than a sequence of changes in the RF channel such as changes in the CQI / RF channel. A training indicator 20, can indicate that the next time the user equipment 110 sends that same identifier, as a predictive identifier, the network node 120 can expect the same sequence of traffic.

[0131] When an indicator 20J is sent to the network node 120 it indicates that each time (i.e. the next time, or until cancelled) that indicator 20J is sent to the network node 120 (to produce an inference), that indicator 20J indicates to the network node 120 that the same sequence of traffic for the user equipment 110 will be expected during a subsequent interval (the observation period 90J). The predictive identifier indicates to the network node 120 that the same corresponding one or more user equipment 110 (software) states was detected and / or sequence of traffic for the user equipment 110 will be expected during a subsequent interval. The training indicator 20 indicates that after training, each time that predictive identifier is sent to the network node 120, the network node 120 should expect / anticipate the same corresponding / mapped network traffic during a corresponding / subsequent interval.

[0132] In some examples, the user equipment 10 determines whether a trigger event based on a user equipment state 10J requires use of an existing indicator 20, re-use of an indicator 20 or a new indicator 20.

[0133] In the following examples, the indicator 20 is a Traffic Pattern Identifier (TPI).

[0134] The UE 110 uses its prior observations of event triggers based on user equipment states 10, and the subsequent data activity to determine that there is a likely known or predictable traffic pattern 60 during the upcoming interval 90 following a particular event trigger based on a particular user equipment state 10.

[0135] For example, if a particular event trigger based on a user equipment state 10 was just detected by the UE 110, and in the past that event trigger was consistently followed by a first pattern of data activity, then the user equipment 110 proceeds to determine what indicator 20 (TPI) it will utilize. First the user equipment 110 checks to see if there is any first TPI 10 that it has previously transmitted to the network, where that previously transmitted first TPI was associated with a particular pattern of data activity which is substantially the same as that first pattern of data activity. If yes then the user equipment 110 can utilize the first TPI associated with the first pattern of data activity. If no, then the user equipment 110 needs to utilize a new TPI. The UE can either utilize a second TPI which has never been transmitted to the network, or can utilize a previously transmitted third TPI, but the user equipment 110 additionally needs to indicate that the prior training for that third TPI needs to be discarded as that third TPI will now serve as an indicator 20 for a new traffic pattern.

[0136] A TPI 20J is used for a trigger event based on a user equipment state 10J that is expected to cause data transfer (UL / DL) within an observation period 90_i. The user equipment manages mapping 30 of trigger events including the user equipment states 10 to TPIs 20.

[0137] If the user equipment 110 has mobility among cells during the observation period 90_i, the UE’s current TPI 20, and the associated TPI observation period 90J can be provided as part of the source destination handoff signaling for the user equipment 110 to the target cell after the handover

[0138] The user equipment 110 / user equipment OS has the opportunity to make multiple observations after a 1st event. As such the user equipment 110 can observe if the timing of the traffic within the latter parts of the observation interval are shifted to be later if earlier packets earlier in that observation interval were delayed while waiting for network (NW) resources / allocations. As such, the user equipment 110 may observe that two different events should utilize two different TPI's because although the two different events sometimes result in the exact same traffic pattern, the first event results in the same traffic pattern of arrivals regardless of any delays in the network provisioning of resources / allocations within that observation interval whereas the second event results in the same traffic pattern of arrivals in the case where the network provides resources with a first level of timeliness, but the second event results in a different traffic pattern of arrivals in the case where the network provides resources after some greater delay, such that traffic within the latter parts of the observation interval are shifted to be later if earlier packets earlier in that observation interval were delayed while waiting for NW resources / allocations.

[0139] The training of a machine learning (ML) model, to provide mappings 50 between respective indicators 20J (TPIs) and respective network traffic predictions 40J, is based on training data. The data collection and ML model training can be performed over / across multiple UEs. The data collection and ML model training can be over / across multiple cells.

[0140] During training, the user equipment 110 can, for example, send a Pattern Status Report (PSR) to the network node 120. A PSR comprises a TPI 20, a TPI Management Field, a TPI Observation Period Start Indicator, a TPI Observation Period Duration Indicator, a TPI Scope Indicator. The PSR can be sent as a single message or as multiple messages.

[0141] The TPI Management Field indicates TPI usage such as clear, merge or copy. The TPI usage ‘clear’ is intended to clear the TPI training previously associated with this TPI e.g. network node should clear any previous mapping associated with that TPI. The TPI usage ‘copy’ is intended to copy and use the TPI training previously associated with an indicated TPI. The training may be accelerated where this TPI can be initialized with the model for another specific TPI.

[0142] The TPI Observation Period Start Indicator comprises a first part indicating whether the observation period 90J begins prior to, or subsequent to the TPI message 20J. The next part indicates the number of slots prior to, or subsequent to the PSR, where the traffic pattern 80_i associated with the TPI 20_i begins (the TPI start slot).

[0143] The TPI Duration Indicator indicates the time duration over which the pattern extends after the TPI start slot.

[0144] The TPI Scope Indicator (e.g. DRB / UE / IMEISV specific or group) indicates that this TPI is (or is not) specific to a UE, for example, using the International Mobile Equipment Identifier (IM El) , or indicates that this TPI is (or is not) shared among a group of UEs, for example, with the same International Mobile Equipment Identifier software version (IMEI SV) or within the same cell or other location or indicates that this TPI is (or is not) for a particular data radio bearer (DRB).

[0145] The preferred embodiment is that Pattern Status Report (PSR) will be sent over the PUSCH as a new MAC CE. The TPI management field, and TPI Scope Indicator could potentially be done over RRC so that it can achieve additional reliability. Medium Access Control (MAC) Control element (CE) or Uplink Control Information (UCI) can be utilized to indicate TPI related information on a more dynamic / per traffic pattern basis.

[0146] During training, the traffic pattern can be based on observations at user equipment 110 (e.g. delay, queueing, and other information not visible to network node 120) and / or based on observations at the network node 120 during the observation period. The observation period can (for training) be before or after the sending of the indicator 20 (TPI). The observation period (for inference) is after the sending of the indicator 20 (TPI). The data traffic pattern can be uplink and / or downlink and / or user plane and / or control plane.

[0147] The data pattern relates to quantity and timing. The timing can be affected by a user equipment 1 10 delay (uplink delay) or a network delay (e.g. UL grant delay). The network can observe whether existing proactive UL grant is under or over used, and / or is too early or just in time. The network can observe if existing DL is underused or user equipment 110 sleeping (DRx) prevents DL.

[0148] The ML model could be a feedforward neural network or a recurrent neural network (e.g. LSTM, transformer).

[0149] Different options can be considered for training data collection. A single user equipment 110 can perform training data collection and then the model is trained and valid for that UE. Training could be realized at the user equipment 110 and used for inference, possibly shared to the network. Training could alternatively be realized at the network.

[0150] Multiple UEs 110 can perform data collection where the network collects the data for multiple UEs and multiple cells. The model training is performed at the network. The model training can be performed at a different network entity than the one where the inference is performed. For example, the model can be trained at the core network and transferred to the edge network for inference. The model training can be performed at the same network entity as the one where the inference is performed. For example, the model can be trained at the base station and inference can be performed at the base station.

[0151] During inference, the traffic pattern (the network traffic prediction 40) is ‘looked-up’ from the trained model (mapping 50) using the received indicator 20.

[0152] During inference, the user equipment 110 can, for example, send a TPI Message comprising the indicator 20_i (TPI). The TPI message can implicitly / explicitly define an ‘observation’ period 90_i e.g. TPI traffic interval reference interval, e.g. (relative) start and end time of traffic pattern associated with that TPI preferably the interval immediately after the TPI.

[0153] The preferred embodiment is that this TPI Message will be sent over the Physical Uplink Shared Channel (PUSCH) as a new MAC CE.

[0154] The network traffic prediction 40 can be used to control resource allocation for the user equipment and control the data transfer pattern used.

[0155] The network traffic prediction 40 can be used to control proactive uplink grants, dynamic / configured grants.

[0156] The network traffic prediction 40 can be used by the NN (120) to control UE awake time (PDCCH monitoring occasions) so as to better manage user equipment 110 active time such that the user equipment 110 is awake when it is most likely that the network will need to provide a DCI over the PDCCH for e.g. downlink traffic arriving for that UE, and / or for a dynamic proactive uplink grant, and sleeping at other times.

[0157] The network traffic prediction 40 can be used to control a variation of downlink control information (DCI) signalling over the physical downlink common control channel (PDCCH) indicating to UEs in a cell (or neighboring cells): resource allocation on the physical downlink shared channel (PDSCH) and / or the physical uplink shared channel (PUSCH)

[0158] UE(s) DRX active time (PDCCH monitoring occasions - PMO) modified to include more or less PMO / active time. The network can avoid excess grants and user equipment 110 awake time when they are not actually needed.

[0159] The network may additionally make other changes to improve power control and / or beam forming to change the coverage / efficiency with which it can cover certain UE(s) where it has now anticipated additional traffic exchanges, or a given type / constraint, based upon the (estimated) TPIs signalled by the user equipment 110 to the network. For example, it could also e.g. impact beamforming, or even the selection of the number of cell Tx / Rx elements, e.g. causing the network to increase the number of network transceiver elements from 32 to 64 if needed in order to carry the traffic pattern(s) now anticipated.

[0160] Feedback to the user equipment 110 on a value of the received indicator 20 to the network can be provided to the user equipment. This can be used to throttle TPI reporting (change regularity).

[0161] The network can observe e.g. via the timing of BSRs (buffer status reports), SRs (scheduling requests), and which (proactive) uplink grants contain user data. These three cases all provide the opportunity for the network to determine if, at a particular point in time, the user has more or less data waiting for transmission on the uplink. A BSR is message which can be sent from the UE to the network, and provides information on data which is currently queuing at the user equipment. An SR is a scheduling request, and under certain conditions, for example where the user equipment has data traffic to exchange, the user equipment will transmit an SR. An uplink grant assigns a portion of the PUSCH (physical uplink shared channel) to the UE such that the user equipment can transmit on that portion of the PUSCH. An uplink grant may be considered proactive if, for example, it is providing a grant for uplink resources even though the most recent BSR report from the UE indicated there is no data pending at the user equipment. When the network decodes the portion of the PUSCH, the network can determine if that allocation contained some amount of user data, or for example only contained padding. As such, these provide the opportunity for the network to determine if, at a particular point in time, the user has more or less data waiting for transmission on the uplink. In the case where the network provides one or more (proactive) uplink grants (CG or DG), the network knows that the user equipment traffic became available for transmission sometime between a previous unused uplink grant, or BSR (e.g. reporting no UE data is pending), or unused SR, and a subsequent used SR, BSR and / or (proactive) uplink grant which carried user data.

[0162] However, the network lacks full visibility into the amount of time that data was queued prior to it being reported and / or carried in a (proactive) uplink grant (and / or reported by a BSR / DSR (Delay Status Report). This information could be useful to the network in determining whether to, in the future after some corresponding TPI, to provide the one or more (additional) proactive uplink grants earlier or not.

[0163] DSR provides info on the packet delay budget remaining for traffic with is queueing in the UE.

[0164] A goodness report can be sent from the user equipment 110 to the network node 120 indicating how long data was queued prior to being serviced by a proactive UL grant. The goodness report be sent separately to the indicator 20 (TPI). The goodness report could be a separate MAC CE message, and could be an extension to e.g. a DSR or BSR.

[0165] Additionally or alternatively, the network can obtain delay information from a delay status report (DSR) received from the user equipment which reports how much of the delay budget remains for a particular set of data, e.g. the network determines how long the data has been queuing by subtracting the amount of delay budget remaining from the total delay budget associated with that data.

[0166] The network node 120 can also be configured to adjust to elasticity. If network traffic is initially delayed then the subsequent pattern may also be time shifted (inelastic) or the subsequent pattern may not be time-shifted (elastic) because the initial delay is compensated for. The ML model can accommodate different elasticity. The number of indicators 20 is limited to reduce the number of bits required to convey the indicator 20. This forces distillation of the context of the user equipment 110 TPI, and reduces computational complexity.

[0167] During learning, the UE may send a 1st indicator, and then send a different indicator 20 or a cancel command. For example, the UE may thus cancel the 1st indicator 20 sent, (pre pattern TPI) and then send a different indicator (post pattern TPI) to overwrite the 1st indicator 20 (the pre pattern TPI). The UE may do this if the traffic pattern the UE observed during the observation period did not match what the UE expected and the UE would prefer to not associate the actual traffic pattern in that observation period with the 1st indicator (pre pattern TPI), and may then prefer to have that traffic pattern be associated with no indicator (cancel) or associated with a different 2nd indicator 20 (the post pattern TPI).

[0168] The network in dependence upon the network traffic predictor 40 obtained using the received indicator 20, can perform subsequent actions / signalling such as: provide grant e.g. configured / dynamic / proactive grants, send updated downlink control information (DCI) e.g. for grants over Physical Downlink Common Control Channel (PDCCH), for adjusting UE power saving (PDCCH skipping, Search Space Set Group (SSSG) switching, UE Discontinuous Reception (DRX), Wake Up signal (WuS) or impacting active time or PDCCH monitoring occasions (PMOs)); cause network entity state changes targeted to be appropriate / ] ust in time for that network traffic pattern.

[0169] Optionally, the network could also then use the network traffic predictor 40 obtained using the received indicator 20, to set up a configured grant, and indicate that a particular configured grant will follow a particular grant pattern thereby avoiding a certain amount of PDCCH signaling.

[0170] Fig 8 illustrates an example of a controller 400 suitable for use in an apparatusl 10, 120. Implementation of a controller 400 may be as controller circuitry. The controller 400 may be implemented in hardware alone, have certain aspects in software including firmware alone or can be a combination of hardware and software (including firmware).

[0171] As illustrated in Fig 8 the controller 400 may be implemented using instructions that enable hardware functionality, for example, by using executable instructions of a computer program 406 in a general-purpose or special-purpose processor 402 that may be stored on a computer readable storage medium (disk, memory etc) to be executed by such a processor 402.

[0172] The processor 402 is configured to read from and write to the memory 404. The processor 402 may also comprise an output interface via which data and / or commands are output by the processor 402 and an input interface via which data and / or commands are input to the processor 402.

[0173] The memory 404 stores a computer program 406 comprising computer program instructions (computer program code) that controls the operation of the apparatus 110, 120 when loaded into the processor 402. The computer program instructions, of the computer program 406, provide the logic and routines that enables the apparatus to perform the methods illustrated in the accompanying Figs. The processor 402 by reading the memory 404 is able to load and execute the computer program 406.

[0174] The apparatus 110 comprises: at least one processor 402; and at least one memory 404 including computer program code, the at least one memory storing instructions that, when executed by the at least one processor 402, cause the apparatus at least to: send to a network node 120 an indicator 20 configured to map a user equipment state 10 to a network traffic prediction 40 for the user equipment 110.

[0175] The apparatus 120 comprises: at least one processor 402; and at least one memory 404 including computer program code, the at least one memory storing instructions that, when executed by the at least one processor 402, cause the apparatus at least to: map a user equipment state 10 to a network traffic prediction 40 for the user equipment 110 based on an indicator 20 received from the user equipment 110.

[0176] As illustrated in Fig 9, the computer program 406 may arrive at the apparatus 110, 120 via any suitable delivery mechanism 408. The delivery mechanism 408 may be, for example, a machine readable medium, a computer-readable medium, a non- transitory computer-readable storage medium, a computer program product, a memory device, a record medium such as a Compact Disc Read-Only Memory (CD- ROM) or a Digital Versatile Disc (DVD) or a solid-state memory, an article of manufacture that comprises or tangibly embodies the computer program 406. The delivery mechanism may be a signal configured to reliably transfer the computer program 406. The apparatus 110, 120 may propagate or transmit the computer program 406 as a computer data signal.

[0177] Computer program instructions for causing an apparatus 110 to perform at least the following or for performing at least the following: send to a network node 120 an indicator 20 configured to map a user equipment state 10 to a network traffic prediction 40 for the user equipment 110.

[0178] Computer program instructions for causing an apparatus 120 to perform at least the following or for performing at least the following: map a user equipment state 10 to a network traffic prediction 40 for the user equipment 110 based on an indicator 20 received from the user equipment 110.

[0179] The computer program instructions may be comprised in a computer program, a non- transitory computer readable medium, a computer program product, a machine readable medium. In some but not necessarily all examples, the computer program instructions may be distributed over more than one computer program.

[0180] Although the memory 404 is illustrated as a single component / circuitry it may be implemented as one or more separate components / circuitry some or all of which may be integrated / removable and / or may provide permanent / semi-permanent / dynamic / cached storage. Although the processor 402 is illustrated as a single component / circuitry it may be implemented as one or more separate components / circuitry some or all of which may be integrated / removable. The processor 402 may be a single core or multi-core processor.

[0181] References to ‘computer-readable storage medium’, ‘computer program product’, ‘tangibly embodied computer program’ etc. or a ‘controller’, ‘computer’, ‘processor’ etc. should be understood to encompass not only computers having different architectures such as single / multi- processor architectures and sequential (Von Neumann) / parallel architectures but also specialized circuits such as field- programmable gate arrays (FPGA), application specific circuits (ASIC), signal processing devices and other processing circuitry. References to computer program, instructions, code etc. should be understood to encompass software for a programmable processor or firmware such as, for example, the programmable content of a hardware device whether instructions for a processor, or configuration settings for a fixed-function device, gate array or programmable logic device etc.

[0182] As used in this application, the term ‘circuitry’ may refer to one or more or all of the following:

[0183] (a) hardware-only circuitry implementations (such as implementations in only analog and / or digital circuitry) and

[0184] (b) combinations of hardware circuits and software, such as (as applicable):

[0185] (i) a combination of analog and / or digital hardware circuit(s) with software / firmware and

[0186] (ii) any portions of hardware processor(s) with software (including digital signal processor(s)), software, and memory or memories that work together to cause an apparatus, such as a mobile phone or server, to perform various functions and

[0187] (c) hardware circuit(s) and or processor(s), such as a microprocessor(s) or a portion of a microprocessor(s), that requires software (for example, firmware) for operation, but the software may not be present when it is not needed for operation.

[0188] This definition of circuitry applies to all uses of this term in this application, including in any claims. As a further example, as used in this application, the term circuitry also covers an implementation of merely a hardware circuit or processor and its (or their) accompanying software and / or firmware. The term circuitry also covers, for example and if applicable to the particular claim element, a baseband integrated circuit for a mobile device or a similar integrated circuit in a server, a cellular network device, or other computing or network device.

[0189] The blocks illustrated in the accompanying Figs may represent steps in a method and / or sections of code in the computer program 406. The illustration of a particular order to the blocks does not necessarily imply that there is a required or preferred order for the blocks and the order and arrangement of the block may be varied. Furthermore, it may be possible for some blocks to be omitted.

[0190] Where a structural feature has been described, it may be replaced by means for performing one or more of the functions of the structural feature whether that function or those functions are explicitly or implicitly described.

[0191] The systems, apparatus, methods and computer programs may use machine learning which can include statistical learning. Machine learning is a field of computer science that gives computers the ability to learn without being explicitly programmed. The computer learns from experience E with respect to some class of tasks T and performance measure P if its performance at tasks in T, as measured by P, improves with experience E. The computer can often learn from prior training data to make predictions on future data. Machine learning includes wholly or partially supervised learning and wholly or partially unsupervised learning. It may enable discrete outputs (for example classification, clustering) and continuous outputs (for example regression). Machine learning may for example be implemented using different approaches such as cost function minimization, artificial neural networks, support vector machines and Bayesian networks for example. Cost function minimization may, for example, be used in linear and polynomial regression and K-means clustering. Artificial neural networks, for example with one or more hidden layers, model complex relationship between input vectors and output vectors. Support vector machines may be used for supervised learning. A Bayesian network is a directed acyclic graph that represents the conditional independence of a number of random variables. As used here ‘module’ refers to a unit or apparatus that excludes certain parts / components that would be added by an end manufacturer or a user.

[0192] The above-described examples find application as enabling components of: automotive systems; telecommunication systems; electronic systems including consumer electronic products; distributed computing systems; media systems for generating or rendering media content including audio, visual and audio visual content and mixed, mediated, virtual and / or augmented reality; personal systems including personal health systems or personal fitness systems; navigation systems; user interfaces also known as human machine interfaces; networks including cellular, non-cellular, and optical networks; ad-hoc networks; the internet; the internet of things; virtualized networks; and related software and services.

[0193] The apparatus can be provided in an electronic device, for example, a mobile terminal, according to an example of the present disclosure. It should be understood, however, that a mobile terminal is merely illustrative of an electronic device that would benefit from examples of implementations of the present disclosure and, therefore, should not be taken to limit the scope of the present disclosure to the same. While in certain implementation examples, the apparatus can be provided in a mobile terminal, other types of electronic devices, such as, but not limited to: mobile communication devices, hand portable electronic devices, wearable computing devices, portable digital assistants (PDAs), pagers, mobile computers, desktop computers, televisions, gaming devices, laptop computers, cameras, video recorders, GPS devices and other types of electronic systems, can readily employ examples of the present disclosure. Furthermore, devices can readily employ examples of the present disclosure regardless of their intent to provide mobility.

[0194] The term ‘comprise’ is used in this document with an inclusive not an exclusive meaning. That is any reference to X comprising Y indicates that X may comprise only one Y or may comprise more than one Y. If it is intended to use ‘comprise’ with an exclusive meaning then it will be made clear in the context by referring to “comprising only one...” or by using “consisting”. In this description, the wording ‘connect’, ‘couple’ and ‘communication’ and their derivatives mean operationally connected / coupled / in communication. It should be appreciated that any number or combination of intervening components can exist (including no intervening components), i.e. , so as to provide direct or indirect connection / coupling / communication. Any such intervening components can include hardware and / or software components.

[0195] As used herein, the term "determine / determining" (and grammatical variants thereof) can include, not least: calculating, computing, processing, deriving, measuring, investigating, identifying, looking up (for example, looking up in a table, a database or another data structure), ascertaining and the like. Also, "determining" can include receiving (for example, receiving information), accessing (for example, accessing data in a memory), obtaining and the like. Also, " determine / determining" can include resolving, selecting, choosing, establishing, and the like.

[0196] In this description, reference has been made to various examples. The description of features or functions in relation to an example indicates that those features or functions are present in that example. The use of the term ‘example’ or ‘for example’ or ‘can’ or ‘may’ in the text denotes, whether explicitly stated or not, that such features or functions are present in at least the described example, whether described as an example or not, and that they can be, but are not necessarily, present in some of or all other examples. Thus ‘example’, ‘for example’, ‘can’ or ‘may’ refers to a particular instance in a class of examples. A property of the instance can be a property of only that instance or a property of the class or a property of a sub-class of the class that includes some but not all of the instances in the class. It is therefore implicitly disclosed that a feature described with reference to one example but not with reference to another example, can where possible be used in that other example as part of a working combination but does not necessarily have to be used in that other example.

[0197] Although examples have been described in the preceding paragraphs with reference to various examples, it should be appreciated that modifications to the examples given can be made without departing from the scope of the claims. Features described in the preceding description may be used in combinations other than the combinations explicitly described above.

[0198] Although functions have been described with reference to certain features, those functions may be performable by other features whether described or not.

[0199] Although features have been described with reference to certain examples, those features may also be present in other examples whether described or not.

[0200] The term ‘a’, ‘an’ or ‘the’ is used in this document with an inclusive not an exclusive meaning. That is any reference to X comprising a / an / the Y indicates that X may comprise only one Y or may comprise more than one Y unless the context clearly indicates the contrary. If it is intended to use ‘a’, ‘an’ or ‘the’ with an exclusive meaning then it will be made clear in the context. In some circumstances the use of ‘at least one’ or ‘one or more’ may be used to emphasis an inclusive meaning but the absence of these terms should not be taken to infer any exclusive meaning.

[0201] The presence of a feature (or combination of features) in a claim is a reference to that feature or (combination of features) itself and also to features that achieve substantially the same technical effect (equivalent features). The equivalent features include, for example, features that are variants and achieve substantially the same result in substantially the same way. The equivalent features include, for example, features that perform substantially the same function, in substantially the same way to achieve substantially the same result.

[0202] In this description, reference has been made to various examples using adjectives or adjectival phrases to describe characteristics of the examples. Such a description of a characteristic in relation to an example indicates that the characteristic is present in some examples exactly as described and is present in other examples substantially as described.

[0203] The above description describes some examples of the present disclosure however those of ordinary skill in the art will be aware of possible alternative structures and method features which offer equivalent functionality to the specific examples of such structures and features described herein above and which for the sake of brevity and clarity have been omitted from the above description. Nonetheless, the above description should be read as implicitly including reference to such alternative structures and method features which provide equivalent functionality unless such alternative structures or method features are explicitly excluded in the above description of the examples of the present disclosure.

[0204] Whilst endeavoring in the foregoing specification to draw attention to those features believed to be of importance it should be understood that the Applicant may seek protection via the claims in respect of any patentable feature or combination of features hereinbefore referred to and / or shown in the drawings whether or not emphasis has been placed thereon. l / we claim:

Claims

CLAIMS1 . A user equipment comprising: means for sending to a network node an indicator configured to map a user equipment state to a network traffic prediction for the user equipment.

2. A user equipment as claimed in claim 1 , comprising: means for sending to the network node at a first time a first indicator configured to map a first user equipment state to a first network traffic prediction for the user equipment; and means for sending to the network node at a second time a second indicator configured to map a second user equipment state to a second network traffic prediction for the user equipment.

3. A user equipment as claimed in claim 1 or 2, comprising: means for receiving a configuration message from the network node that configures at the user equipment sending of the indicator.

4. A user equipment as claimed in claim 3, wherein the configuration message is a radio resource control (RRC) message or wherein the configuration message is a medium access control (MAC) message.

5. A user equipment as claimed in claim 3 or 4, wherein the configuration message comprises one or more information elements configured to: configure the indicator; and / or configure an observation period associated with the indicator and / or configure the user equipment state and / or configure at least one trigger condition for sending the indicator and / or configure a likelihood of sending the indicator configured to map a user equipment state to a network traffic prediction for the user equipment.

6. A user equipment as claimed in claim 3, 4 or 5, wherein the configuration message comprises one or more information elements configured to: configure a plurality of indicators including a first indicator and a second indicator.

7. A user equipment as claimed in claim 6, wherein the configuration message comprises one or more information elements configured to: configure one or more observation periods associated with the plurality of indicators including the first indicator and the second indicator; and / or configure multiple user equipment states including the first user equipment state and / or the second user equipment state; configure trigger conditions for sending an indicator configured to map a user equipment state to a network traffic prediction including a first trigger condition for sending the first indicator and a second trigger condition for sending the second indicator.

8. A user equipment as claimed in any preceding claim, wherein the indictor is dependent upon an identifier of the user equipment previously transferred to the network or wherein the first indictor, but not the second indicator, is dependent upon an identifier of the user equipment previously transferred to the network.

9. A user equipment as claimed in any preceding claim, comprising means for receiving a configuration message from the network node, that configures the user equipment to: stop sending any indicator or stop sending a specified indicator or change an observation period associated with an indicator or change a state associated with an indicator or change or prevent a trigger condition causing sending of an indicator.

10. A user equipment as claimed in any preceding claim wherein the user equipment state is a state of the user equipment expected to cause data transfer within an observation period, and / or wherein the user equipment state is at least one of a state of an operating system or traffic prediction system of the user equipment or a state of an application running on the user equipment.11 . A user equipment as claimed in any preceding claim, wherein the user equipment state is triggered by a change in state of the operating system of the user equipment or a change in state of an application running on the user equipment, and / or wherein the user equipment state is triggered by an action of a user of the user equipment.

12. A user equipment as claimed in any preceding claim, wherein the user equipment state is a hash of application parameters for an application running on the user equipment, wherein the application is capable of transferring data between the user equipment and the network node.

13. A user equipment as claimed in any preceding claim, comprising means for sending to the network node information on delays at the user equipment in association with the indicator.

14. A user equipment as claimed in any preceding claim, configured to enter a learning mode configured by the network mode to create a mapping that maps, directly or indirectly, from the indicator to the network traffic prediction for the user equipment.

15. A user equipment as claimed in any preceding claim, comprising: means for sending to the network node the indicator before a mapping, enabling the indicator to map the user equipment state to the network traffic prediction for the user equipment, has been established, to enable establishment of the mapping; means for sending to the network node the indicator configured to map the user equipment state to the network traffic prediction for the user equipment after the mapping has been established.

16. A user equipment as claimed in any preceding claim, comprising means for receiving a configuration message from the network node that configures at the user equipment machine learning model training and / or means for receiving a configuration message from the network node that configures at the user equipment inference using a machine learning model at the network node.

17. A user equipment as claimed in any preceding claim, comprising means for: detecting that a trigger conditions for sending an indicator is met; in response to detecting that a trigger condition for sending an indicator has been met, causing sending to the network node the indicator configured to map the user equipment state to the network traffic prediction for the user equipment.

18. A computer program that when run on one or more processors of a user equipment, causes the user equipment to send to a network node an indicator configured to map a user equipment state to a network traffic prediction for the user equipment.

19. A method comprising: sending to a network node an indicator configured to map a user equipment state to a network traffic prediction for the user equipment.

20. A network apparatus comprising: means for receiving from a user equipment an indicator configured to map a user equipment state to a network traffic prediction for the user equipment.21 . A network apparatus as claimed in claim 20, configured to create a mapping that maps, directly or indirectly, from the received indicator to the network traffic prediction for the user equipment.

22. A network apparatus as claimed in claim 20 or 21 , comprising: means for receiving from the user equipment before a mapping, enabling the indicator to map the user equipment state to the network traffic prediction for the user equipment, has been established, to enable establishment of the mapping; means for receiving from the user equipment the indicator configured to map the user equipment state to the network traffic prediction for the user equipment after the mapping has been established.

23. A network apparatus as claimed in claim 20 or 21 , comprising: means for sending a configuration message to the user equipment that configures the user equipment to assist with machine learning model training at the network apparatus, wherein thetrained machine learning model enables mapping an indicator received from the user equipment to a network traffic prediction for the user equipment.

24. A computer program that when run on one or more processors of a network apparatus, causes the network apparatus to map a user equipment state to a network traffic prediction for the user equipment based on an indicator received from the user equipment.

25. A method comprising: receiving from a user equipment an indicator configured to map a user equipment state to a network traffic prediction for the user equipment; mapping a user equipment state to a network traffic prediction for the user equipment based on the received indicator.

Images