Neural Bit Loading and Rate Control for Spectral Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing wireless communication systems face inefficiencies in bit loading and channel coding rate control, leading to suboptimal performance and increased complexity due to the use of uniform constellations and coding rates across channel conditions, which can impact spectral efficiency and signaling overhead.
Innovation Solution
Implementing a deep-learning based neural network (NN) model for dynamic bit loading and channel coding rate control, allowing for adaptive selection of constellations and coding rates based on channel conditions, reducing complexity and optimizing receiver performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If fixed constellations and channel code rates are used across all subblocks of a transport block, then device complexity is reduced, but spectral efficiency and communication performance deteriorate due to inability to adapt to varying channel conditions
Solution Approach 1:
The transport block is divided into multiple subblocks, and each subblock is independently assigned different constellations and channel coding rates based on channel conditions. This segmentation allows the system to adapt to varying channel conditions across different parts of the transmission without requiring the entire block to use complex adaptive schemes, thus improving spectral efficiency while keeping individual subblock processing manageable.
Solution Approach 2:
The patent implements dynamic selection of constellations and channel coding rates for each subblock based on real-time channel conditions. Instead of using fixed parameters throughout the transport block, the system dynamically adjusts modulation and coding parameters to match the instantaneous channel state, thereby optimizing spectral efficiency without requiring the receiver to handle excessively complex adaptive processing.
2Productivity
If dynamic bit loading and channel coding rate selection is implemented for each resource element, then spectral efficiency is improved, but device complexity and signaling overhead increase
Solution Approach 1:
The system segments the transport block into subblocks and applies dynamic parameter selection at the subblock level rather than at the individual resource element level. This segmentation reduces the granularity of complexity while still capturing the benefits of adaptation to channel variations, thereby improving spectral efficiency without overwhelming the transmitter and receiver with excessive processing complexity.
Solution Approach 2:
Different constellations and channel coding rates are applied to different subblocks based on their specific channel conditions. This local optimization allows each subblock to be processed with parameters tailored to its local channel characteristics, improving overall spectral efficiency while distributing the computational complexity across multiple simpler subblock operations rather than requiring complex global optimization.
3Reliability
If per-subblock constellation and coding rate control is implemented, then communication performance is improved, but signaling overhead and latency increase
Solution Approach 1:
The system performs preliminary selection of constellations and channel coding rates for subblocks based on channel condition predictions or pre-established mapping rules. By preparing these parameter selections in advance based on predicted or historical channel conditions, the system can quickly apply appropriate parameters without requiring extensive real-time signaling exchanges, thereby improving communication performance while minimizing signaling latency.
Solution Approach 2:
The patent introduces an intermediary mechanism where the transmitter and receiver share a pre-established mapping between channel condition indicators and parameter selections. This intermediary mapping table allows both sides to independently determine the appropriate constellations and coding rates without requiring continuous explicit signaling, thus improving communication performance while reducing the signaling overhead and latency associated with real-time parameter negotiation.
Data Source
AI summary
Apparatuses and methods for ML based dynamic bit loading and rate control are described. An apparatus obtains NN model coefficient information associated with communication channel information, between a UE and network, and the UE capabilities. The apparatus generates, using the NN model and based on the coefficient information, a communication configuration including a constellation or a CB channel coding rate. The apparatus communicates, with the network and based on the CB of a TB, using the NN-generated channel coding rate or constellation. Another apparatus obtains NN model coefficient information, associated with communication channel information, between a UE and network, and the UE capabilities. The apparatus generates, using the NN model and based on the coefficient information, a communication configuration including a constellation or a CB channel coding rate. The apparatus communicates, with the UE and based on the CB of a TB, using the NN-generated channel coding rate or constellation.