PIM Signal Estimation Using Neural Network Evaluation Model
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for estimating the impact of passive intermodulation (PIM) signals in communication technologies are inadequate, particularly when PIM signals overlap with receive frequency bands, leading to inaccurate impact assessments.
Innovation Solution
A signal processing method that involves scheduling bandwidth resources, determining characteristic data from downlink signal data, and using a signal strength evaluation model to accurately estimate PIM signal strength across multiple cells, thereby assessing the impact of PIM signals on each cell.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a current estimation method is used where transmit sub-bands are sequentially selected and frequency bands are determined that do not overlap with PIM signals, then the frequency band determination is simple, but the PIM signal impact cannot be accurately estimated when PIM signals overlap with the receive frequency band
Solution Approach 1:
The patent introduces a machine learning model as an intermediary between raw signal data and PIM impact estimation. The model takes multiple signal characteristics (RSSI, SINR, RSRP, RSRQ) as inputs and processes them through trained neural networks to produce accurate PIM strength estimates. This intermediary structure enables complex non-linear relationships to be captured without requiring explicit complex processing algorithms.
Solution Approach 2:
The patent transforms the estimation approach by changing from direct frequency band overlap analysis to using multiple signal parameters (RSSI, SINR, RSRP, RSRQ) as inputs to a machine learning model. This parameter transformation allows the system to capture subtle PIM effects that are not apparent through simple frequency overlap detection, significantly improving estimation accuracy.
2Productivity
If bandwidth resources are increased to accommodate more cells and frequency bands, then the system capacity improves, but the PIM interference problem becomes more prominent
Solution Approach 1:
The patent implements a feedback mechanism where the machine learning model continuously estimates PIM signal strength for each cell based on real-time signal characteristics. This feedback information is then used to identify and avoid frequency bands with high PIM interference, allowing the system to maintain high capacity while dynamically adapting to PIM conditions. The feedback loop enables proactive interference management rather than reactive mitigation.
Data Source
Figure 1
Figure 2(a)
Figure 2(b)
AI summary
Embodiments of this application disclose a signal processing method and a related apparatus, to estimate PIM signal strength. The method in embodiments of this application includes: First downlink signal data corresponding to each of N cells in a preset bandwidth resource scheduling solution is obtained. Then, first characteristic data corresponding to each piece of the first downlink signal data is determined. Then, first passive intermodulation PIM signal strength corresponding to a first cell is determined based on a preset signal strength evaluation model and the first characteristic data corresponding to each of the N cells, to accurately estimate impact of the first PIM signal on the first cell.