Sub-Nyquist Signal Compression Using Delay-Line Phase Compensation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional data compression techniques using compressed sensing have a high computational load, making them impractical for processing broadband signals at rates above 1 GS/s.
Innovation Solution
A data compression apparatus that branches a target signal into multiple lines, applies different delay times, and samples at a sub-Nyquist rate, followed by frequency-domain conversion and phase compensation to determine the sub-Nyquist zone and estimate the target signal's frequency, converting it into compressed data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If compressed sensing technique is used to represent sparse signal with fewer variables, then compression ratio is improved, but computational load increases making processing impractical for broadband signals at 1 GS/s or more
Solution Approach 1:
The patent segments the frequency spectrum into multiple sub-Nyquist zones and processes signals in parallel through multiple channels with different delay times. This segmentation allows the system to handle broadband signals at high speeds by dividing the complex optimization problem into smaller, parallelizable tasks, thereby maintaining compression ratio while improving processing throughput to accommodate 1 GS/s or more
Solution Approach 2:
The patent performs preliminary actions by pre-defining multiple sub-Nyquist zones and their corresponding basis functions before signal processing. The system pre-calculates and stores transformation matrices for each zone, so that during actual signal processing, the encoder can quickly map sampled values to the appropriate zone and reconstruct the signal without performing complex real-time optimization, thus reducing computational load while maintaining compression efficiency
Data Source
AI summary
A data compression apparatus includes a receiver that outputs sampling sequences corresponding to signals obtained by adding different delay times to different signals obtained by branching a target signal into a plurality of lines, and sampling the signals at a sampling rate less than the Nyquist rate, and an encoder that converts the sampling sequences into compressed data. The encoder includes a sub-FFT that converts the sampling sequences into frequency-domain signals, a signal processing unit that performs, at one time, phase compensation processing for sub-Nyquist zones of the sampling sequences, and processing to cancel phase rotation, a target frequency estimator that determines into which sub-Nyquist zone the target signal has been folded and estimates the frequency of the target signal, and an encoding unit that converts a value representing the sub-Nyquist zone and the corresponding amplitude value into the compressed data and output the compressed data.