GFDM Filter Coefficient Adjustment for Demodulation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In Generalized Frequency Division Multiplexing (GFDM), the demodulation process on the reception side is restricted by the resource setting, particularly when the number of subsymbols and subcarriers in a unit resource are both even, leading to difficulties in zero-forcing demodulation and increased bit error rates when the number of subsymbols is even.
Innovation Solution
The solution involves setting filter coefficients based on different rules depending on whether the number of subsymbols in a unit resource is odd or even, ensuring the transformation matrix is regular and invertible, and applying an offset corresponding to half the sample frequency to minimize interference between subcarriers, thereby enabling successful demodulation even when the number of subsymbols is odd.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the number of subsymbols and subcarriers in a unit resource are both even, then the resource setting flexibility is improved, but the demodulation success is worsened due to transformation matrix singularity
Solution Approach 1:
The patent applies parameter changes by dynamically adjusting the filter coefficient sampling position based on the parity of the number of subsymbols. When the number of subsymbols is even, the sampling position is offset by half a sample frequency relative to the default position. This parameter adjustment ensures the transformation matrix remains regular and invertible, resolving the demodulation failure issue while maintaining resource setting flexibility.
2Adaptability or versatility
If the number of subsymbols is even, then the resource allocation flexibility is improved, but the bit error rate increases due to interference between subcarriers
Solution Approach 1:
The patent applies local quality by implementing different filter coefficient sampling strategies for different cases (odd vs. even number of subsymbols). Specifically, when the number of subsymbols is even, the sampling position is locally adjusted by an offset of half a sample frequency. This localized adjustment minimizes interference between subcarriers and reduces bit error rates while preserving the overall flexibility of resource allocation.
3Reliability
If the transformation matrix is made regular by changing filter coefficient sampling position, then the demodulation capability is improved, but the frequency characteristic of the pulse shaping filter changes
Solution Approach 1:
The patent applies dynamics by making the filter coefficient sampling position adjustable based on the number of subsymbols. The sampling position is not fixed but dynamically changed: for even numbers of subsymbols, an offset of half a sample frequency is applied, while for odd numbers, the default position is used. This dynamic adjustment ensures the transformation matrix remains regular for demodulation while adapting the frequency characteristics to maintain optimal performance.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
[Object] To provide a mechanism that can make a success of demodulation on the reception side even if the restrictions on the resource setting are lifted. [Solution] An apparatus including: a processing unit configured to variably set at least any of a number of subcarriers or a number of subsymbols included in a unit resource including one or more subcarriers or one or more subsymbols, and set a filter coefficient that is applied in accordance with different rules which depend on whether the number of subsymbols included in the unit resource is an even number or an odd number.