Granted Patent B2

US 12,542,837 · App. 18/559,114 · Granted Feb 3, 2026

Inventors: Dario Bega (Munich, DE); Anna Pantelidou (Massy, FR); Christiane Maria Allwang (Munich, DE)

Assignee: Nokia Technologies Oy

H04L67/62H04L41/147H04L41/16

Abstract

Claims (51)

1 . A requests prediction apparatus implemented as a Network Data Analytics Function (NWDAF) in a 5G telecommunication network, the requests prediction apparatus comprising:

at least one processor; and

at least one memory storing instructions that, when executed by the at least one processor, cause the requests prediction apparatus at least to perform:

receiving, from a requests prediction service client, a request for a requests prediction service, wherein the request for comprises:

one or more probability thresholds, and

one or more future time intervals;

upon receiving the request, retrieving, from a requests storage entity implemented as an analytics data repository function, information related to one or more past data service requests comprising network data including channel state measurements, quality of service measurements, and user equipment measurements;

training an artificial intelligence or a machine learning algorithm based on the retrieved information;

generating information related to one or more predicted data service requests based on the trained artificial intelligence or machine learning algorithm, wherein the information related to one or more predicted data service requests includes a probability value for each predicted data service request of the one or more predicted data service requests, wherein the probability value represents a probability of receiving the predicted data service request at a future time interval of the one or more future time intervals, and wherein each of the one or more predicted data service requests comprises:

an indication of a requesting network function expected to send the predicted data service request, and

a type of the predicted data service request;

selecting, from the one or more predicted data service requests, at least one predicted data service request comprising: an associated probability value higher than a probability threshold of the one or more probability thresholds, and a predicted data value exceeding a data value threshold; and

sending, to the requests prediction service client, a response to the request for the requests prediction service, wherein the response comprises the at least one selected predicted data service request together with trigger information for enabling the prediction service client to perform proactive data collection or proactive training of an artificial intelligence model prior to receiving the predicted data service request.

2 . The requests prediction apparatus of claim 1 , wherein

the one or more past data service requests correspond to one or more requests received, at one or more previous time intervals, by the requests prediction service client, and

the one or more predicted data service requests correspond to one or more requests expected to be received, at the future time interval, by the requests prediction service client.

3 . The requests prediction apparatus of claim 2 , wherein the at least one memory and the instructions, when executed by the at least one processor, further cause the requests prediction apparatus at least to perform:

generating one or more predicted values based on data obtained using the one or more past data service requests;

determining whether one or more trigger criteria are met based on the one or more predicted values; and

sending, to the requests prediction service client, trigger information including information related to the data for which the one or more trigger criteria are met.

4 . The requests prediction apparatus of claim 3 , wherein the information related to the at least one predicted data service request further comprises a parameter of the at least one predicted data service request.

5 . The requests prediction apparatus according to claim 4 , wherein the information related to each of the one or more past data service requests comprises:

an indication of the requesting entity that has sent the past data service request,

an indication on a past time interval during which the past data service request was received,

an indication on the type of the past data service request, and

a parameter of the past data service request.

6 . A prediction service client apparatus implemented as a Network Function (NF) in a 5G telecommunication network, the prediction service client apparatus comprising:

at least one processor; and

at least one memory storing instructions that, when executed by the at least one processor, cause the prediction service client apparatus at least to perform:

sending, to a requests predictor implemented as a network data analytics function, a request for a requests prediction service, wherein the request comprises:

a unique identifier of the network function;

one or more probability thresholds, and

one or more future time intervals;

receiving, from the requests predictor, a response to the request for the requests prediction service, the response comprising information related to at least one predicted data service request, wherein the information related to the at least one predicted data service request comprises at least one of:

an indication of a requesting entity expected to send the predicted data service request;

a type of the predicted data service request; and

a probability value associated with the predicted data service request;

triggering, based on trigger information including the information related to the at least one predicted data service request, proactive actions comprising:

proactive data collection of telecom network data comprising channel state measurements, quality of service measurements, and user equipment measurements from one or more data sources, and

proactive training of an artificial intelligence or machine learning model using the proactively collected telecom network data, prior to receiving the predicted data service request; and

wherein the triggering information indicates that a probability value and a predicted data value exceed respective thresholds.

7 . The prediction service client apparatus of claim 6 , wherein the at least one memory and the instructions, when executed by the at least one processor, further cause the prediction service client apparatus at least to perform:

receiving, from the requests predictor, the trigger information indicating that one or more trigger criteria based on predicted values are met, wherein the predicted values are generated based on values of data obtained using the one or more past data service requests.

8 . The prediction service client apparatus of claim 7 , wherein the information related to the at least one predicted data service request further comprises:

a probability value associated with the at least one predicted data service request, and

a parameter of the at least one predicted data service request.

9 . The prediction service client apparatus of claim 6 , wherein information related to each of the past data service requests comprises:

an indication of the requesting entity that has sent the past data service request,

an indication on a past time interval during which the past data service request was received,

an indication on the type of the past data service request, and

a parameter of the past data service request.

Assignments (4)

Continuity (1)

References Cited (29)