#parametric-estimation

Dew.Signal.Core

Dew.Signal.Core is the cross-platform, managed-only edition of the Dew.Signal DSP library. It provides a comprehensive suite of digital signal processing algorithms and components without requiring native binary acceleration. This maximizes compatibility across platforms including Windows, Linux, macOS, mobile, embedded, and WebAssembly environments. Filter Design and Processing: - IIR filters: Butterworth, Chebyshev I/II, Elliptic, Bessel; analog and discrete design workflows - Frequency transformations in Z- and S-domains, bilinear and matched-Z transforms, pole/zero and state-space forms, group delay analysis, stability filters - FIR filter design: window methods, Remez exchange algorithm, Hilbert transformers, differentiators, integrators, Savitzky–Golay filtering, fast envelope detectors - Multi-rate filtering: half-band multistage FIR designs, zoom-spectrum, decimation, interpolation Nonlinear Processing: - Sample-and-hold, sample-and-decay, and median filtering Spectral Analysis and Estimation: - FFT-based spectral analysis framework - Parametric estimators: Yule–Walker, Burg, Covariance, Modified Covariance - CZT, real and complex cepstrum and inverse cepstrum - Bispectrum, bicoherence, coherence and transfer function estimation - Peak detection, peak interpolation and tracking, phase unwrapping - Advanced spectral statistics: noise floor, SFDR, THD, THDN, SINAD, RMS, SNR Time-Domain Modeling and System Representation: - Conversion between state-space, zero-pole, numerator–denominator transfer models - System zero/pole determination and stability analysis Signal Synthesis, Streaming, and Measurement: - Signal generators using stack-based vectorized function evaluation - Support for multiple noise models: white, pink, brownian, blue, violet - Components for continuous streaming, recording, playback, triggering and monitoring - Scalable FIR and IIR convolution, auto-correlation, cross-correlation, DCT/IDCT Forecasting and Time-Series Modeling: - Spectral forecasting based on user-defined peak selection for interpretable predictive modeling Platform and Integration Notes: - Uses Dew.Math.Core as its numerical foundation (managed vector/matrix computing) - Visualization is optionally provided through Dew.Signal.Tee (Windows-only), not included here - Compatible with cross-platform UI frameworks (MAUI, Uno, Avalonia), compute services, and WASM Dew.Signal.Core is ideal when portability, sandbox safety, or plugin isolation is required.