Neural Network Radio Resource Management for Adaptive UE Scanning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional radio resource management (RRM) in wireless networks relies on static or algorithmic configurations for user equipment (UE), which fail to account for the UE's particular circumstances, leading to inefficient information acquisition and excessive power consumption.
Innovation Solution
Implementing a neural-network-based RRM scheme that fuses sensor data and radio measurements to dynamically adapt RRM actions, using a trained neural network (NN) to optimize UE operations based on current context.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a static or algorithmic configuration is used for UE RRM operations, then the network can manage RRM at a large-scale level with simplified control, but the UE cannot adapt to its particular circumstances, resulting in non-optimal RRM behavior and excessive power consumption
Solution Approach 1:
The patent transforms the static RRM configuration into a dynamic system by implementing a neural network that continuously adapts RRM parameters based on real-time sensor data and radio measurements. The neural network processes multiple input features (sensor readings, radio measurements, RRM actions) and dynamically adjusts configuration parameters, enabling the UE to adapt to changing environmental conditions and operational contexts while maintaining manageable complexity through the learned policy.
2Ease of manufacture
If a static RRM configuration is employed, then implementation and deployment are straightforward, but the UE performs unnecessary RRM actions that consume excessive power and fail to account for specific UE circumstances
Solution Approach 1:
The patent dynamically changes RRM configuration parameters based on the UE's operational context by using a neural network to process sensor data and radio measurements. The system adjusts parameters such as measurement frequencies, scanning intervals, and handover thresholds according to the UE's movement state, environmental conditions, and network conditions, thereby optimizing power consumption while maintaining RRM effectiveness.
3Ease of operation
If a static schedule is used for intra-frequency, inter-frequency, or inter-RAT scanning, then the network infrastructure can direct UEs with uniform configuration, but the scanning does not account for the UE's particular circumstances, leading to inefficient information acquisition
Solution Approach 1:
The patent implements preliminary action by using sensor data (accelerometer, gyroscope, magnetometer) and radio measurements to predict the UE's movement and environmental context before performing RRM scanning actions. The neural network processes these inputs to determine the optimal scanning schedule in advance, adjusting intra-frequency, inter-frequency, and inter-RAT scanning based on predicted UE behavior and network conditions, thereby improving information acquisition efficiency while maintaining ease of operation through centralized neural network management.
Data Source
AI summary
A wireless device is configured to receive a set of sensor data from one or more sensors of the wireless device, receive a set of radio measurements from a radio interface of the wireless device, process the set of sensor data and the set of radio measurements at a radio resource management (RRM) neural network of the wireless device to generate an output representative of an RRM action, and then perform the RRM action. The wireless device further can provide a representation of sensor capabilities of the wireless device for receipt by an infrastructure component of a network infrastructure that is wirelessly connected to the wireless device, receive a neural network architectural configuration from the infrastructure component in response to providing the representation of sensor capabilities, and implement the neural network architectural configuration at the RRM neural network.