Beam Index Prediction for Wireless Device Power Reduction

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current beam management procedures in wireless communications networks require inefficient resource utilization, as wireless devices scan and measure multiple beams to identify the best beam, leading to unnecessary power consumption and resource waste.

Innovation Solution

A method using a wireless device with a controller and a beam predictor, trained on tuples associating geographic positions and times with beam indices, to predict and report the best beam index to the access network node, reducing the need for extensive measurements and resource usage.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If the wireless device scans and measures multiple beams to identify the best beam, then the beam selection accuracy is improved, but the power consumption and resource usage increase

Engineering Contradiction:

Improvebeam selection accuracyVSAvoidpower consumption

Core Design Contradiction:

Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The system performs preliminary beam training during idle periods to establish beam associations with geographic positions before actual data transmission. The beam predictor is pre-trained with tuples containing geographic positions, times, and beam indices, so that during data transmission the device can directly query the predicted beam index without performing full beam sweeping, thereby avoiding unnecessary power consumption while maintaining accurate beam selection

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

Instead of directly measuring all beams during data transmission, the system creates a copy of beam measurement data collected during idle periods and stores it in the beam predictor. This copied data is then used to predict the best beam during data transmission, eliminating the need for repeated full beam measurements and reducing power consumption while preserving beam selection accuracy

Inventive Principle:

Principle #26Copying

2Reliability

If the wireless device reports multiple beam indices, then the network can select the best beam, but the uplink resource usage and reporting overhead increase

Engineering Contradiction:

Improvebeam selection reliabilityVSAvoiduplink resource usage

Core Design Contradiction:

ReliabilityVSQuantity of substance

Solution Approach 1:

The system extracts only the essential beam index information from the beam prediction result and reports only this single extracted parameter to the network during data transmission. The beam predictor internally processes the full beam measurement data during idle periods to identify the single best beam index, so that during data transmission only this one extracted index needs to be reported, significantly reducing uplink resource usage while maintaining reliable beam selection

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The beam predictor performs preliminary processing of beam measurement data during idle periods to pre-identify the single best beam index before data transmission. This preliminary action eliminates the need to report multiple beam indices during data transmission, as the best beam is already determined in advance, thereby reducing reporting overhead and uplink resource usage while maintaining selection reliability

Inventive Principle:

Principle #10Preliminary action

3Adaptability or versatility

If the network transmits reference signals in multiple beams, then the wireless device can find the best beam, but the downlink resource usage and network power consumption increase

Engineering Contradiction:

Improvebeam coverageVSAvoidnetwork power consumption

Core Design Contradiction:

Adaptability or versatilityVSUse of energy by stationary object

Solution Approach 1:

The network performs preliminary beam sweeping during idle periods to collect beam measurement data and train the beam predictor with tuples of geographic positions, times, and beam indices. During data transmission, the network uses the predicted beam index to transmit reference signals only in the predicted best beam direction, eliminating the need for repeated full beam sweeping and significantly reducing downlink resource usage and network power consumption while maintaining adaptability through the pre-collected beam data

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system discards the practice of transmitting reference signals in all beams during data transmission and recovers the beam measurement data collected during idle periods by using it to train the beam predictor. This allows the network to discard unnecessary downlink transmissions during data transmission while recovering the useful information from idle period measurements, thereby reducing network power consumption while maintaining beam coverage adaptability

Inventive Principle:

Principle #34Discarding and recovering

Data Source

PatentUS20230412235A1Reporting of a beam index in a communications network

Publication Date: 2023.12.21 TELEFONAKTIEBOLAGET LM ERICSSON (PUBL)

AI summary

A method for reporting a beam index by a wireless device including a controller, a beam predictor, and a radio transceiver. The method incudes obtaining an indication that the wireless device is to report a beam index to an access network node and obtaining the beam index from the beam predictor by providing information of current geographic position of the wireless device and current time as input to the beam predictor. The beam predictor is trained with tuples that associate at least different geographical positions and different points in time with different beam indices. The beam predictor predicts the beam index based on the current geographic position and the current time. The method includes wirelessly reporting, using the radio transceiver, the beam index as obtained from the beam predictor to the access network node.