Block-Based Crest Factor Reduction for Low-Overhead Peak Cancellation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing Crest Factor Reduction techniques process signals sample-by-sample, leading to inefficiencies due to software processing overhead, which impairs the efficiency of power amplifiers in wireless communication systems.
Innovation Solution
Implementing block-based Crest Factor Reduction methods that process data in blocks of samples, using a vector engine and incorporating pre-cursor and post-cursor blocks to ensure continuity and reduce edge effects, with iterative processing to address peak regrowth.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If sample-by-sample Crest Factor Reduction processing is used, then peak cancellation can be performed, but processing efficiency deteriorates due to software overhead
Solution Approach 1:
The patent divides the continuous signal processing into discrete blocks of samples. Each block is processed independently through the CFR algorithm, allowing batch processing optimizations. This segmentation enables the system to amortize software overhead across multiple samples within each block, significantly improving processing efficiency while maintaining effective peak cancellation performance.
2Productivity
If block-based processing is implemented, then processing efficiency improves, but edge effects and discontinuities are introduced
Solution Approach 1:
The patent applies pre-cursor blocks before the main data block and post-cursor blocks after it. These cursor blocks are processed along with the main block to ensure that peak cancellation operations at block boundaries do not cause discontinuities. The pre-cursor blocks allow peaks near the beginning of the main block to be properly canceled, while post-cursor blocks handle peaks that extend beyond the main block boundaries, thereby maintaining signal continuity and stability.
3Manufacturing precision
If peak cancellation is applied aggressively, then PAR reduction improves, but noise introduction increases
Solution Approach 1:
The patent applies peak cancellation selectively based on local signal characteristics. The algorithm identifies actual peaks exceeding a threshold and applies cancellation only at those locations, rather than uniformly across the entire signal. This localized approach ensures effective PAR reduction at peak locations while minimizing noise introduction in regions where peaks are absent, thereby improving the signal-to-noise ratio overall.
Solution Approach 2:
The patent employs iterative processing where the output of one CFR pass becomes the input for subsequent passes. In each iteration, newly emerged peaks that were not present or were below threshold in previous iterations are detected and canceled. This feedback mechanism ensures progressive PAR reduction while allowing the system to adapt to the changing signal characteristics, preventing excessive noise introduction by stopping when convergence is achieved.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Block-based crest factor reduction (CFR) techniques are provided. An exemplary block-based crest factor reduction method comprises obtaining a block of data samples comprised of a plurality of samples; applying the block of data to a crest factor reduction block; and providing a processed block of data from the crest factor reduction block. The block-based crest factor reduction method can optionally be iteratively performed a plurality of times for the block of data. The block of data samples can comprise an expanded block having at least one cursor block. For example, at least two pre-cursor blocks and one post-cursor block can be employed. The peaks can be cancelled, for example, only in the block of data samples and in a first of the pre-cursor blocks.