Dew.Signal.Linux is the Linux-native accelerated edition of the Dew.Signal digital signal processing library. It provides a full suite of DSP algorithms built on top of Dew.Math.Linux, delivering high-performance numerical processing with multithreaded AVX/AVX2/AVX-512 hardware acceleration. This package is intended for scientific servers, compute clusters, HPC pipelines, digital instrumentation, real-time data acquisition, industrial analytics, embedded Linux platforms, and cloud CPU workloads. Filter Design and Processing: - IIR filters: Butterworth, Chebyshev I/II, Elliptic, Bessel - Transformations: bilinear, matched-Z, frequency remapping, pole-zero and state-space formulations - FIR filters: window methods, Remez exchange, Hilbert transformers, differentiators, integrators, Savitzky–Golay smoothers, envelope detection - Multirate DSP: decimation, interpolation, half-band polyphase filters, zoom-spectrum workflows Spectral and Frequency-Domain Analysis: - FFT-based spectral estimation and spectrum analyzer infrastructure - Parametric estimators: Yule–Walker, Burg, Covariance, Modified Covariance - Chirp-Z transform, time-frequency spectrograms, bispectrum, bicoherence, coherence, transfer function estimation, phase unwrapping - Real/complex cepstrum and inverse cepstrum - Spectral statistics: noise floor, SFDR, THD, THDN, SINAD, RMS, SNR Signal Modeling, Streaming, and Synthesis: - White, pink, brownian, blue, violet and deterministic test signal generators - Continuous streaming components and dataflow processing units for real-time measurement systems - High-performance convolution, correlation, DCT/IDCT, interpolation and filtering kernels - Spectral forecasting based on controlled peak selection Integration and Platform Model: - Uses Dew.Math.Linux for native-accelerated numerical backend - Does **not** require WinForms or TeeChart (visualization is optional and external) - Suitable for server, embedded, batch compute, containerized, and headless execution Dew.Signal.Linux provides the same API surface as Dew.Signal, but is optimized specifically for Linux compute environments where high throughput and deterministic performance are required.
Get Started
$ dotnet add package Dew.Signal.LinuxReadme
Dew.Signal.Linux
High-performance Digital Signal Processing (DSP) for Linux
Dew.Signal.Linux is the Linux-native accelerated edition of the Dew.Signal DSP library. It provides a full suite of signal processing algorithms built on top of Dew.Math.Linux, using multithreaded AVX/AVX2/AVX-512 hardware acceleration for high-throughput numerical processing.
This package is designed for scientific servers, compute clusters, HPC pipelines, digital instrumentation, real-time data acquisition systems, industrial analytics, embedded Linux platforms, and cloud CPU workloads.
Key Features
Filter Design and Processing
- IIR filters: Butterworth, Chebyshev I/II, Elliptic, Bessel
- Analog and digital filter workflows
- Frequency transformations in S and Z domains
- Bilinear and Matched-Z transforms
- Group delay and stability analysis
- State-space, zero-pole, and numerator-denominator transfer function modeling
- FIR filter design: window methods and Remez exchange algorithm
- Hilbert transformers, differentiators, integrators, Savitzky-Golay smoothing, envelope detectors
- Multi-rate DSP: half-band multistage FIR design, decimation, interpolation, zoom-spectrum workflows
Spectral and Frequency-Domain Analysis
- FFT-based spectral analysis and streaming spectrum analyzer infrastructure
- Parametric estimators: Yule-Walker, Burg, Covariance, Modified-Covariance
- Chirp-Z transform and time-frequency spectrogram computation
- Bispectrum, bicoherence, coherence and transfer function estimation
- Phase unwrapping and peak interpolation
- Real and complex cepstrum and inverse cepstrum
- Spectral statistics: noise floor, SFDR, THD, THDN, SINAD, RMS, SNR
Signal Modeling, Streaming, and Synthesis
- Signal generators for deterministic, white, pink, brownian, blue, and violet noise
- Components for continuous streaming, playback, logging, triggering, and measurement
- High-performance convolution and correlation
- 1D and 2D filtering, DCT and inverse DCT