NN Symbol Mapping Across Multiple Resource Elements
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current neural network-based models for wireless communications suffer from complexity and issues related to beamforming, channel estimation, and spectral efficiency, particularly when dealing with channel impairments such as noise and fading, as they typically map each input bit to a single symbol, leading to information loss when the symbol is not detected.
Innovation Solution
Introduce an NN-based symbol generator within the transmission chain that allows one data bit to be associated with multiple REs, incorporating additional inputs like channel quality metrics, enabling adaptive symbol generation compatible with conventional precoding and beamforming techniques.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If each input bit is mapped to a single symbol in neural network-based models, then the model structure remains simple, but information loss occurs when the symbol is not detected and spectral efficiency deteriorates
Solution Approach 1:
The patent segments the mapping relationship by dividing the set of allocated resource elements into multiple groups, where each group is associated with different input bits. This allows one input bit to be mapped to multiple symbols across different groups, reducing information loss while maintaining manageable model structure through structured segmentation.
Solution Approach 2:
The patent introduces an additional dimension to the mapping relationship by allowing bits to map to multiple symbols across different resource element groups rather than a single symbol. This dimensional expansion enables redundancy and improves spectral efficiency without proportionally increasing model complexity.
2Measurement precision
If dedicated resource elements are allocated for reference signals, then channel estimation accuracy improves, but spectral efficiency decreases due to reduced data transmission capacity
Solution Approach 1:
The patent makes certain resource elements serve multiple functions: they act as reference signals for channel estimation and simultaneously carry data information. By training the neural network to handle both reference signal processing and data transmission, the system achieves accurate channel estimation without sacrificing spectral efficiency, as the same resource elements fulfill both roles.
Solution Approach 2:
The patent changes the functional parameters of resource elements dynamically. Instead of dedicating fixed resource elements solely to reference signals, the system adjusts the role of resource elements based on channel conditions and training outcomes, allowing flexible allocation that optimizes both channel estimation accuracy and spectral efficiency.
3Reliability
If conventional neural network models are used without adaptive channel quality inputs, then the model structure remains simple, but resilience to channel impairments such as noise and fading deteriorates
Solution Approach 1:
The patent incorporates channel quality metrics as inputs to the neural network, creating a feedback mechanism where the model adapts its symbol generation based on current channel conditions. This allows the system to respond dynamically to noise and fading, improving reliability while the modular architecture keeps the complexity increase manageable.
Solution Approach 2:
The patent introduces dynamic adaptation by allowing the neural network to adjust symbol generation based on varying channel quality metrics. Instead of a static mapping, the system dynamically modifies its behavior according to current channel conditions, enhancing resilience to impairments while maintaining a relatively simple base model structure.
Data Source
AI summary
Various aspects of the present disclosure relate to neural network (NN)-based techniques that support the association of one data bit (e.g., input bit) to multiple resource elements (REs) via NN-based mapping blocks (or NN-based symbol generators) within a transmission chain of a transmitting node. For example, the transmission chain may insert the NN-based mapping block between a layer mapping block and precoding block, enabling the NN-based mapping block to generate output symbols from a sequence of modulation symbols. The NN-based mapping block, therefore, may operate to generate some or all RE symbols for each spatial layer, enabling adaptive symbol generation while maintaining compatibility within a transmission chain.