Adaptive LLR Quantization for Fronthaul Bandwidth Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The separation of demodulation and decoding functions in cellular networks leads to high data flow rates in the fronthaul network, exceeding the capacity of communication buses, particularly when LLR data are transmitted between separate devices, causing inefficiencies and potential signal degradation.
Innovation Solution
A method for quantifying demodulated signals using adaptive scalar quantization tables based on channel coding levels, reducing the data size by optimizing quantization intervals according to the distribution of LLR values, and implementing inverse quantization for efficient channel decoding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If demodulation and decoding functions are separated into different devices, then functional flexibility and centralized management are improved, but data flow rate requirements exceed communication bus capacity
Solution Approach 1:
The patent extracts only the essential information from the full LLR data by selecting a subset of LLR values based on quantization thresholds, rather than transmitting all LLR data. This reduces the data volume transmitted between demodulation and decoding functions while preserving the most significant information for channel decoding.
Solution Approach 2:
The patent changes the representation parameters of LLR data by applying quantization with specific thresholds to transform continuous LLR values into discrete quantized levels. This parameter transformation significantly reduces the bandwidth required for transmitting soft bits between separated demodulation and decoding functions.
2Reliability
If full precision LLR data are transmitted, then decoding performance is maintained, but bandwidth requirements exceed fronthaul network capacity
Solution Approach 1:
The patent applies different quantization thresholds to different portions of the LLR distribution, concentrating representation on the most probable LLR values while using coarser quantization for less probable values. This local differentiation maintains decoding performance for critical values while reducing overall bandwidth requirements.
Solution Approach 2:
The patent transmits only a partial representation of the full LLR data by selecting and quantizing specific LLR values based on predetermined thresholds, rather than transmitting all possible LLR values. This partial transmission is sufficient for achieving good decoding performance while dramatically reducing bandwidth consumption.
3Quantity of substance
If quantization levels are reduced, then bandwidth consumption decreases, but signal precision and decoding accuracy deteriorate
Solution Approach 1:
The patent performs preliminary selection of LLR values to be quantized based on predetermined thresholds before transmission. By pre-identifying which LLR values are most important to preserve, the system can use fewer quantization levels while maintaining signal precision for the critical values that most impact decoding accuracy.
Data Source
Figure 1
Figure 2A
Figure 2B
AI summary
The invention relates to a method for quantising data representative of a radio signal received by a radio antenna of a mobile network, comprising: - demodulation of the radio signal received by the antenna, providing a demodulated signal, - scalar quantisation of each value of the demodulated signal using a quantisation table selected according to a channel coding level used to transmit the radio signal, providing a quantised demodulated signal, - transmission of the quantised demodulated signal to a channel decoding module.