Method and apparatus for CBRS network planning and operating in an enterprise network

Patent metadata

Field	Value
Publication number	US-11997505-B2
Application number	US-202017010584-A
Country	US
Kind code	B2
Filing date	Sep 2, 2020
Priority date	Sep 2, 2020
Publication date	May 28, 2024
Grant date	May 28, 2024

Abstract

A method and apparatus for planning a wireless communication network operating in a spectrum-controlled radio band and determining spectrum availability at an enterprise location. A plurality of network models are generated using machine learning techniques, the network models are provided to a network planner unit, input is received regarding intended deployment, and spectrum availability is determined responsive to the enterprise location. The number of BS/APs needed is estimated, and an enterprise network may be planned and deployed. The enterprise network is monitored, and if network operation is not meeting expected performance, the models may be retrained and the network planner updated. Spectrum availability is determined by registering a discovery group with an SAS at a location to provide spectrum availability, and repeating for each location. The method may be periodically performed to generate a time series of data which may be analyzed to provide spectrum availability.

First claim

What is claimed is:

1. A method of planning, deploying, and operating a spectrum-controlled, wireless communication network that includes a plurality of Base Stations/Access Points (BS/APs) located within an enterprise location, the BS/APs communicating wirelessly with a plurality of User Equipment devices (UEs), comprising the steps of:
generating a plurality of network models using machine learning techniques responsive to a predictor space that includes at least one of spectrum availability, topology definitions, Node/Link configurations, pathloss models, and line-of-sight probabilities;
storing said network models in a models database;
providing said network models to a network planner;
receiving input into said network planner regarding intended deployment, including location and at least one of building type, network requirements, and number of devices;
responsive to said location input, determining spectrum availability at said location; and
responsive to said input and spectrum availability, estimating and deploying a number of BS/APs needed to cover the intended deployment.

2. The method of claim 1 wherein the method further comprises the steps of:
providing a predictor space that includes at least one of spectrum availability, topology definitions, Node/Link configurations, pathloss models, and line-of-sight probabilities;
performing a Discrete Event/Monte Carlo simulation responsive to said predictor space;
storing successful iterations of said simulation in a Predictor Database; and
generating said network models responsive to said successful iterations.

3. The method of claim 1 further comprising the steps of:
planning a network responsive to said estimate, and deploying the plan for an enterprise network;
monitoring operation of said enterprise network, including collecting statistics regarding network operation; and
determining whether or not said operation is meeting key performance indicators (KPIs);
if it is meeting the KPIs then continuing network operation without change,
if it is not meeting the KPIs then retraining said model and updating the network planner with the retrained model.

4. The method of claim 1 wherein the method further comprises the steps of:
predicting spectrum availability at a plurality of locations; and
selecting one of said locations responsive to said location input, and providing the associated predicted spectrum availability to the network planner.

5. The method of claim 4 wherein the spectrum is controlled by a Spectrum Access System (SAS), and said step of predicting spectrum availability includes the steps of inquiring regarding spectrum availability with said SAS at each of said plurality of locations, and storing said spectrum availability for each of said plurality of locations.

6. The method of claim 4 further comprising the steps of:
modifying said network plan responsive to said predicted spectrum availability; and
deploying said wireless network responsive to said modified network plan.

7. A network orchestration module for planning and administering a plurality of spectrum-controlled, wireless communication networks, each network including a plurality of Base Stations/Access Points (BS/APs) located at an enterprise location, the BS/APs communicating wirelessly with a plurality of User Equipment devices (UEs), the network orchestration module comprising:
a model generator module including a network training module that includes machine learning and artificial intelligence for generating a plurality of network models responsive to a predictor database;
a models database for storing the network models;
a Spectrum Availability Repository that stores spectrum availability at a plurality of locations;
a network planner connected to the models database for storing and implementing at least one of said network models, said network planner also connected to the Spectrum Availability Repository;
means for receiving input into said network planner regarding intended deployment of an enterprise network, said input including enterprise network location; and
wherein the network planner further is connected to the Spectrum Availability Repository for determining spectrum availability responsive to the enterprise network location.

8. The network orchestration module of claim 7 further comprising:
a Spectrum Inquiry Service Module for predicting spectrum availability at each of a plurality of locations and storing said predicted spectrum availability in the Spectrum Availability Repository.

9. The network orchestration module of claim 8 wherein said Spectrum Inquiry Service Module is connected to a remote Spectrum Access Service (SAS), and includes means for inquiring with the SAS regarding spectrum availability at a plurality of locations.

10. The network orchestration module of claim 9 further comprising means for periodically repeating said inquiry with the SAS to provide a time series of spectrum availability data for each of the plurality of locations.

11. The network orchestration module of claim 7 wherein:
said input to the network planner further includes at least one of building type, network requirements, and number of devices; and
further comprising an estimation module, responsive to said input and spectrum availability, for estimating a number of BS/APs needed to cover the enterprise deployment, and means for planning a network responsive to said inputs and said estimate of number of BS/APs.

12. The network orchestration module of claim 11 further comprising means for planning a network responsive to said inputs and said estimate of number of BS/APs.

13. The network orchestration module of claim 7 wherein the plurality of spectrum-controlled networks, include Citizen's Broadband Radio Service (CBRS) networks, and the BS/APs comprise CBRS devices (CBSDs).

14. The network orchestration module of claim 7 further comprising:
a Discrete Event/Monte Carlo Simulation Module including
a predictor space input that includes at least one of spectrum availability, topology definitions, Node/Link configurations, pathloss models, and line-of-sight probabilities;
means for performing a Discrete Event/Monte Carlo simulation responsive to said predictor space; and
a Predictor Database for storing successful iterations of said simulation;

wherein the Predictor Database is connected to the Model Generator Module, which generates at least one network model responsive to said successful iterations.

15. The network orchestration module of claim 7 , wherein the network training module further includes a retraining unit that includes machine learning and artificial intelligence for retraining the plurality of network models, and further comprising:
means for receiving data regarding network operation; and
means responsive to said data for determining whether or not said network operation is meeting key performance indicators (KPIs), if it is meeting KPIs then continuing network operation without change, if it is not meeting KPIs then retraining said model and updating the network planner with the retrained model.

16. The network orchestration module of claim 7 wherein said means for receiving input into said network planner includes a network planner UI.

Assignees

• Celona Inc

Inventors

• Mahalingam Nagi
• Goyal Mohit
• Yavuz Mehmet
• Ponna Rajeshwar
• Shah Rajeev
• Bandyopadhyay Sourav
• Jaiswal Aparna

Classifications

• H04W16/18 Primary

  Network planning tools · CPC title

• G06F30/27

  using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model · CPC title

• G06N20/00

  Machine learning · CPC title

• H04B17/3913

  Predictive models, e.g. based on neural network models · CPC title

• G06F11/3692

  for test results analysis · CPC title