Dew.Signal.Linux
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.
v6.3.10↙ 9.6K