# US 2023/0068437 A1 · Using user-side contextual factors to predict cellular radio throughput

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2023/0068437 A1
- Document ID: US20230068437A1
- Status: Published application
- Kind: A1
- Application number: 17/820,833
- Filed: 2022-08-18
- Published: 2023-03-02
- Earliest claimed: 2021-08-18
- Assignees: Regents of the University of Minnesota
- Applicants: Regents of the University of Minnesota
- Inventors: Arvind Narayanan; Eman Ramadan; Feng Qian; Zhi-Li Zhang
- Web page: https://patentgrep.com/patent/US20230068437A1

## CPC

- H04W 24/02: Arrangements for optimising operational condition
- H04W 24/08: Testing, {supervising or monitoring} using real traffic
- G06N 3/0442: characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
- G06N 3/0455: Auto-encoder networks; Encoder-decoder networks
- G06N 5/01: Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
- G06N 5/045: Explanation of inference; Explainable artificial intelligence [XAI]; Interpretable artificial intelligence
- G06N 20/20: Ensemble learning
- H04W: WIRELESS COMMUNICATION NETWORKS (broadcast communication H04H; communication systems using wireless links for non-selective communication, e.g. wireless extensions H04M1/72)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

A system and method for predicting one or more cellular performance parameters associated with user equipment (UE) within a three-dimensional (3D) space having one or more cellular nodes, the cellular nodes including one or more cellular nodes, including a 5G cellular node. For each of one or more of pieces of UE within the 3D space, determine values associated with one or more UE-side features of each piece of UE. Predict values of the one or more cellular performance parameters for each UE as a function of the values associated with the one or more UE-side features of each respective piece of UE, wherein predicting values of the one or more cellular performance parameters includes applying the values determined for each respective piece of UE to a machine learning module trained using truth data associated with the one or more UE-side features.

## Claims

### Claim 1

A method for predicting one or more cellular performance parameters associated with user equipment (UE) within a three-dimensional (3D) space having one or more cellular nodes, the cellular nodes including one or more cellular nodes, including a 5G cellular node, the method comprising:
determining, for each of one or more of pieces of UE within the 3D space, values associated with one or more UE-side features of each piece of UE; and
predicting values of the one or more cellular performance parameters for each UE as a function of the values associated with the one or more UE-side features of each respective piece of UE,
wherein predicting values of the one or more cellular performance parameters includes applying the values determined for each respective piece of UE to a machine learning module trained using truth data associated with the one or more UE-side features.

### Claim 2 (depends on 1)

The method of claim 1, wherein the cellular performance parameters include one or more of cellular data throughput, signal strength and level of carrier aggregation.

### Claim 3 (depends on 2)

The method of claim 2, wherein cellular data throughput is one or more of downlink throughput and uplink throughput.

### Claim 4 (depends on 1)

The method of claim 1, wherein the method further comprises:
selecting network channels based on the predicted values of the one or more cellular performance parameters to optimize one or more of cellular data throughput, quality of service, handoff, and data prefetch across the one or more pieces of UE.

### Claim 5 (depends on 1)

The method of claim 1, wherein the method further comprises:
predicting cellular data performance across the one or more pieces of UE based on the predicted values of the one or more cellular performance parameters.

### Claim 6 (depends on 1)

The method of claim 1, wherein the UE-side features include mobility features, the mobility features including direction and speed of movement relative to the 3D space for each piece of UE.

### Claim 7 (depends on 6)

The method of claim 6, wherein at least one of the pieces of user equipment is not moving within the 3D space.

### Claim 8 (depends on 1)

The method of claim 1, wherein the UE-side features include connection-based features.

### Claim 9 (depends on 8)

The method of claim 8, wherein the UE-side features further include tower-based features, wherein the tower-based features are selected from features including distance between panel and UE, UE-panel positional angle and UE-panel mobility angle.

### Claim 10 (depends on 1)

The method of claim 1, wherein the UE-side features include tower-based features.

### Claim 11 (depends on 10)

The method of claim 10, wherein the tower-based features are selected from features including distance between panel and UE, UE-panel positional angle and UE-panel mobility angle.

### Claim 12 (depends on 11)

The method of claim 11, wherein the UE-side features further include mobility features.

### Claim 13 (depends on 12)

The method of claim 12, wherein the UE-side features further include connection-based features, the connection-based features including past values of cellular data throughput associated with the UE.

### Claim 14 (depends on 1)

The method of claim 1, wherein the UE-side features include factors, associated with each respective piece of UE, that attenuate 5G signals.

### Claim 15 (depends on 1)

The method of claim 1, wherein the UE-side features include mobility features, the mobility features including direction and speed of movement relative to the 3D space for each piece of UE.

### Claim 16 (depends on 1)

The method of claim 1, wherein the pieces of UE include one or more of cellular telephones, computer tablets and computers.

### Claim 17

A system comprising:
one or more cellular nodes, including one or more 5G panels; and
a computing system connected to the cellular nodes, the computing system including a machine learning module,
wherein the computing system is configured to determine, for each of one or more pieces of user equipment (UE) within a 3D space surrounding the plurality of cellular nodes, values associated with one or more UE-side features of each piece of UE, and
wherein the machine learning module is trained to predict values of one or more cellular performance parameters for each piece of UE as a function of the values associated with the one or more UE-side features of each respective piece of UE, wherein predicting values of the one or more cellular performance parameters includes applying the values determined for each respective piece of UE to the machine learning module after the machine learning module has been trained using truth data associated with the one or more UE-side features.

### Claim 18 (depends on 17)

The system of claim 17, wherein the cellular performance parameters include one or more of cellular data throughput, signal strength and level of carrier aggregation.

### Claim 19 (depends on 18)

The system of claim 18, wherein the computing system comprises a laptop, a server, or a cloud-computing platform.

### Claim 20

A non-transitory, computer-readable medium comprising executable instructions; which when executed by processing circuitry, cause a computing device to:
determine, for each of one or more of pieces of UE within a 3D space, values associated with one or more UE-side features of each piece of UE; and
predict values of the one or more cellular performance parameters for each UE as a function of the values associated with the one or more UE-side features of each respective piece of UE, wherein predicting values of the one or more cellular performance parameters includes applying the values determined for each respective piece of UE to a machine learning module trained using truth data associated with the one or more UE-side features.

### Claim 21

A method for predicting cellular performance for user equipment (UE) within a three-dimensional (3D) space having a plurality of cellular nodes, the cellular nodes including two or more 5G panels and at least one 4G tower, the method comprising:
estimating 4G cellular performance for each UE;
determining, for each of a plurality of pieces of UE within the 3D space, values associated with one or more UE-side features of each piece of UE;
estimating 5G cellular performance for each UE as a function of the values, wherein estimating cellular performance includes applying the values to a machine learning module trained using truth data associated with the one or more UE-side features; and
determining, for each piece of UE and based on the estimated 4G cellular data performance and the estimated 5G cellular data performance, a combination of 4G data traffic and 5G data traffic needed to optimize cellular performance across the plurality of pieces of UE.

### Claim 22 (depends on 21)

The method of claim 21, wherein cellular performance includes optimizing one or more of channel selection, cellular data throughput, quality of service, handoff, and data prefetch.
