Cell Shaping with Reinforced Learning for Antenna Phase Offset
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current wireless communication systems, such as LTE and NR, face challenges in cell shaping due to the lack of phase calibration in radios with fewer branches, leading to impaired coverage and the need for additional cells to compensate for random coverage shapes.
Innovation Solution
The implementation of reinforced learning methods, specifically using a Thompson sampling algorithm, to determine optimal phase offsets for antenna elements, thereby adjusting the common weight for 4-branch radios to achieve better cell coverage and shape.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If phase calibration is not performed in radios with fewer branches, then hardware cost and software complexity are reduced, but coverage shape becomes random and impairment occurs
Solution Approach 1:
The patent changes the phase offset parameters dynamically through reinforced learning. Instead of fixed phase values, the system learns optimal phase offsets (ΔA, ΔB) that adapt to different cell geometries and user distributions, transforming the static parameter into a learnable variable that optimizes coverage without requiring hardware phase calibration
Solution Approach 2:
The system implements feedback mechanisms where user equipment reports measurements (RSRP, SINR) that are used to update the phase offset decisions. The network node receives feedback about coverage quality and adjusts phase offsets accordingly, creating a closed-loop system that continuously optimizes cell shape based on actual performance
2Device complexity
If a fixed common beamforming weight is used for all cells, then device complexity is reduced, but coverage optimization for different user distributions becomes impossible
Solution Approach 1:
The system employs self-service through automated reinforced learning algorithms that automatically determine optimal phase offsets without manual configuration. The network node independently learns and adapts phase settings based on feedback from user equipment, eliminating the need for complex manual optimization while achieving adaptability to different cell scenarios
Solution Approach 2:
The patent introduces dynamics by making phase offsets adjustable and adaptive rather than fixed. The system can dynamically change phase values based on learned patterns and feedback, allowing the same hardware to optimize for different user distributions and cell geometries without requiring complex per-cell configuration
3Reliability
If only one branch from each polarization is used for common weight, then unknown common beam direction is mitigated, but coverage is reduced due to 50% power utilization
Solution Approach 1:
The system changes the phase parameters of all four branches simultaneously in a coordinated manner. By adjusting phase offsets (ΔA, ΔB) across all branches rather than selecting only one branch per polarization, the system maintains full power utilization while achieving beam direction control through constructive interference of signals from all antenna elements
Data Source
AI summary
A method and network node for cell shaping with reinforced learning are disclosed. In some embodiments. for each of a plurality of phase offset trial values: a first reward in response to first trial value applied to antenna elements having a first polarization is determined. A second reward in response to a second trial value applied to antenna elements having a second polarization is determined. A first phase offset apply to the antenna element having the first polarization is determined based at least in part on the plurality of first rewards and on a probable reward in response to the first phase offset. A second phase offset to be applied to antenna elements having the second polarization is determined based at least in part on the plurality of second rewards and on a probable reward in response to the second phase offset.