ML Radio Receiver with Time-Frequency Processing for Distorted Signals
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing digital radio receivers struggle with efficiently processing nonlinearly distorted signals and supporting multiple frequency-multiplexed client devices without requiring ML processing before the FFT block, leading to hardware complexity and inefficiency.
Innovation Solution
A machine learning (ML) model-based radio receiver that incorporates inverse fast Fourier transform (IFFT) and fast Fourier transform (FFT) blocks within its frequency-domain processing, allowing it to alternate between time and frequency domains, effectively separating frequency-multiplexed client devices and compensating for nonlinear distortions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If ML processing is performed before the FFT block, then detection accuracy for nonlinearly distorted signals is improved, but hardware complexity and implementation difficulty increase
Solution Approach 1:
The patent inverts the conventional processing order by performing ML processing after the FFT block rather than before it. This allows the ML model to operate on frequency-domain representations directly, avoiding the need for time-domain ML processing before FFT, thereby reducing hardware complexity while maintaining detection accuracy for nonlinearly distorted signals.
Solution Approach 2:
The patent transforms the signal into the frequency domain using FFT before applying ML processing. This dimensional transformation enables the ML model to process signals in a different domain (frequency rather than time), which simplifies the hardware implementation while preserving the ability to detect nonlinear distortions.
2Device complexity
If conventional receiver processing is used, then hardware implementation is simpler, but performance under heavy distortion and high mobility conditions deteriorates
Solution Approach 1:
The patent changes the domain parameter from time-domain to frequency-domain processing by incorporating FFT blocks before the ML model. This parameter change enables the system to maintain simple hardware implementation while achieving superior performance under heavy distortion and high mobility conditions through frequency-domain signal representation.
3Adaptability or versatility
If ML processing is applied to frequency-multiplexed signals, then support for multiple client devices is improved, but processing time and computational complexity increase
Solution Approach 1:
The patent segments the frequency-multiplexed signals into separate frequency domains for different client devices using FFT-based decomposition. This segmentation allows the ML model to process each client's signal independently in the frequency domain, enabling multi-device support while reducing overall processing time through parallelization.
Data Source
AI summary
Radio receiver devices and related methods and computer programs are disclosed. A radio signal comprising information bits is received at a radio receiver device. The radio receiver device deter-mines log-likelihood ratios, LLRs, of the information bits. The determining of the LLRs comprises applying a machine learning (ML) model to a frequency domain representation of the received radio signal. The ML model is executable to process the frequency domain representation of the received radio signal and to output estimates of the LLRs based on results of the processing. The ML model comprises a first frequency domain processing block, an inverse fast Fourier transform (IFFT) block subsequent to the first frequency domain processing block, a time domain processing block subsequent to the IFFT block, and a fast Fourier transform (FFT) block subsequent to the time domain processing block.