Back to results
US 12,696,135 B2Granted
Traffic balancing for moving users, proactive slice management, and predictive slice management using artificial intelligence
AssigneeBoost SubscriberCo L.L.C.
InventorsDilip Tandekar, Gurpreet Sohi, Sourabh Gupta
Life of this patent
Estimating the term…
Abstract
Traffic balancing for moving users, proactive slice management, and predictive slice management using Artificial Intelligence (AI) are disclosed. Radio Access Networks (RANs) are adjusted based on predicted data usage for moving users based on the applications they are running using AI/machine learning (ML) models. Load balancing is performed when new users move into a coverage area. Predictive slice management is performed to allocate bandwidth to users for a period of time based on predicted application usage.
Claims18
One or more computing systems, comprising:
- memory storing computer program instructions;
- and
- at least one processor configured to execute the computer program instructions, wherein the computer program instructions are configured to cause the at least one processor to:
- receive traffic information from one or more base stations of a Radio Access Network (RAN),
- determine, by a near-real time RAN Intelligent Controller (RT RIC) that uses one or more Artificial Intelligence (AI)/Machine Learning (ML) models, based on the received traffic information that one or more network slices for a mobile device should be configured,
- send, by the RT RIC, control instructions to at least one base station of the one or more base stations to configure the one or more network slices for the mobile device;
- store the traffic information, user profile information, signal information, and spectrum information for the RAN in a data lake;
- train, by a non-real time (NRT) RIC or another application of a network core, the one or more AI/ML models using the stored traffic information, the user profile information, the signal information, and the spectrum information from the data lake;
- and
- deploy, by the NRT RIC, the one or more trained AI/MVL models to the RT RIC via an AI interface.
7 more elements
A computer-implemented method, comprising:
- monitoring, by a near-real time Radio Access Network (RAN) Intelligent Controller (RT RIC), network availability information and load information from a RAN;
- determining, by the RT RIC using one or more Artificial Intelligence (AI)/Machine Learning (ML) models, that proactive load balancing should be performed for the RAN upon one or more User Equipment (UE) devices entering a coverage area;
- and
- sending, by the RIC, control instructions to one or more base stations of the RAN to perform the proactive load balancing;
- and
- storing the traffic information, user profile information, signal information, and spectrum information for the RAN in a data lake;
- training, by a non-real time (NRT) RIC or another application of a network core, the one or more AI/NL models using the stored traffic information, the user profile information, the signal information, and the spectrum information from the data lake;
- and
- deploying, by the NRT RIC, the one or more trained AI/ML models to the RT RIC via an AI interface.
6 more elements
One or more computing systems, comprising:
- memory storing computer program instructions;
- and
- at least one processor configured to execute the stored computer program instructions, wherein the computer program instructions are configured to cause the at least one processor to:
- monitor, by a near-real time Radio Access Network (RAN) Intelligent Controller (RT RIC), a RAN using one or more Artificial Intelligence (AI)/Machine Learning (ML) models,
- determine, by the RT RIC using the one or more AI/ML models, that predictive slice management should be performed based on predicted application usage on a User Equipment (UE) device,
- configure and allocate, by the RT RIC, one or more network slices to the UE device based on the predicted application usage;
- retrieve, by a non-real time (NRT) RIC, signal information, traffic information, network load information, user profile information, and spectrum information from a data lake;
- train, by the NRT RIC, the one or more AI/ML models using the signal information, the traffic information, the network load information, the user profile information, and the spectrum information retrieved from the data lake;
- and
- deploy, by the NRT RIC, the trained one or more AI/ML models to the RT RIC.
7 more elements
Dependencies are inferred from the claim text.