PDLUT Compression Through Symmetry and Folding for Lower-Power Hardware
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The increasing size of Pattern Dependent Lookup Tables (PDLUT) for higher modulation formats leads to complex hardware implementation and increased power consumption in communication systems, particularly for high-order modulation formats like 64QAM.
Innovation Solution
A method and system for compressing PDLUTs by exploiting symmetry and unification properties, involving rearrangement, folding, and amplitude reversal of data symbol sequences and distortion correction values, followed by approximate value determination and clustering using PCA-based AI techniques to reduce the PDLUT size.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a full PDLUT is used for higher modulation formats, then distortion compensation performance is improved, but hardware complexity and power consumption increase
Solution Approach 1:
The patent segments the complete PDLUT into multiple sub-PDLUTs, each handling a specific portion of the distortion correction task. This segmentation allows the system to process symbol sequences in smaller, more manageable units, reducing the hardware complexity of storing and accessing the full PDLUT while maintaining the overall distortion compensation performance through coordinated use of multiple sub-tables
Solution Approach 2:
The patent implements a nested structure where multiple levels of PDLUT are organized hierarchically. The first PDLUT processes initial symbol sequences, and subsequent PDLUTs process corrected sequences in a nested manner. This nesting allows the system to achieve comprehensive distortion compensation through multiple passes while keeping each individual PDLUT smaller and less complex than a single full-size table
2Reliability
If a full PDLUT is used for higher modulation formats, then distortion compensation performance is improved, but power consumption increases
Solution Approach 1:
By segmenting the PDLUT into sub-PDLUTs, the system reduces the power consumption associated with memory access and processing. Each sub-PDLUT requires less energy to access and process compared to a single large PDLUT, while the cumulative effect of multiple sub-PDLUTs maintains the overall distortion compensation performance
Solution Approach 2:
The patent applies partial action by using multiple smaller PDLUTs in sequence rather than one comprehensive PDLUT. Each PDLUT performs a partial correction on the symbol sequence, and the cumulative effect of these partial actions achieves the desired overall distortion compensation while consuming less power than a single full-size lookup operation
3Device complexity
If PDLUT size is reduced through compression, then hardware complexity is reduced, but distortion compensation accuracy may deteriorate
Solution Approach 1:
The patent applies preliminary action by performing distortion correction in multiple sequential passes through different PDLUTs. The first PDLUT performs an initial correction on the symbol sequence, and subsequent PDLUTs perform additional corrections on the already-corrected sequence. This preliminary action in multiple stages maintains high accuracy while allowing each individual PDLUT to be smaller and less complex
Data Source
AI summary
The disclosed systems, and methods are directed to compressing a Pattern Dependent Look-up Table (PDLUT) including distortion correction values corresponding to data symbol sequences, including: i) symmetrically rearranging the data symbol sequences and the associated distortion correction values; ii) folding and amplitude reversing a second half of the rearranged data symbol sequences and the associated distortion correction values; iii) determining approximate distortion correction values for a first half of the rearranged data symbol sequences based on the distortion correction values associated with the first half of the rearranged data symbol sequences and the folded and amplitude reversed distortion correction values associated with the second half of the data symbol sequences; and iv) updating the PDLUT based on the approximate distortion values corresponding to the first half of the rearranged data symbol sequences.