asv-sdr

asv-sdr is a set of .NET libraries for software-defined radio (SDR), IQ stream processing, digital filters, ADS-B/Mode-S parsing, simulation, and selected SDR hardware integrations.

The core package provides a fluent, observable-style processing pipeline for IQ samples. Device-specific projects add bindings and helpers for supported SDR hardware.

Documentation

Read the Asv.Sdr documentation for the project overview, installation, development, and documentation publishing instructions.

Features

• Fluent IQ stream processing based on IReaderIq<T> and IReaderIqSubject<T>.
• DSP helpers for magnitude, phase, differential phase, FFT, frequency shift, overlap, windowing, I/Q manipulation, and custom transforms.
• Reusable filters: moving average, median, Kalman, biquad, sinc/CIC, elliptic, pulse correlation, and window filters.
• ADS-B and Mode-S frame detection, normalization, parsing, and message models.
• AM/FM helpers for extracting modulation, DDM/SDM, frequency offset, and code ID.
• Simulation source for ILS-like IQ signals.
• Hardware integration projects for LimeSDR, Signal Hound, MSI SDR, EVSG, GSPN1, and AD936x/libiio-based SDR devices.

Projects

Project	Purpose
Asv.Sdr	Core DSP, fluent IQ processing, filters, math helpers, ADS-B/Mode-S parser.
Asv.Sdr.Simulate	Virtual IQ sources for tests and offline processing experiments.
Asv.Sdr.LimeSdr	LimeSDR reader/writer and protocol-specific helpers.
Asv.Sdr.SignalHound	Signal Hound BB/SA/SG API bindings and helpers.
Asv.Sdr.Msi	MSI/SDRplay-compatible native API wrapper.
Asv.Sdr.Evsg	EVSG device integration.
Asv.Sdr.Gspn1	GSPN1 serial device integration.
Asv.Sdr.AdSdr	AD936x/libiio-based SDR integration.
Asv.Sdr.Test	xUnit test project.

Requirements

• .NET SDK with support for the target frameworks used by the solution.
• The current shared target framework is configured in src/Directory.Build.props as net10.0.
• Some device projects require native vendor libraries to be installed and discoverable through the runtime library path.

Native library expectations include:

• LimeSDR: LimeSuite
• Signal Hound: bb_api.dll, sa_api.dll, sg_api.dll
• MSI/SDRplay-compatible wrapper: mirsdrapi-rsp
• AD936x/libiio project: libiio-sharp.dll

Installation

Use the package that matches the layer you need:

dotnet add package Asv.Sdr
dotnet add package Asv.Sdr.Simulate
dotnet add package Asv.Sdr.LimeSdr
dotnet add package Asv.Sdr.SignalHound

For local development, reference the projects from src/Asv.Sdr.slnx.

Build

From the repository root:

dotnet restore src/Asv.Sdr.slnx
dotnet build src/Asv.Sdr.slnx -c Release

Run tests:

dotnet test --solution src/Asv.Sdr.slnx -c Release

Create NuGet packages:

dotnet pack src/Asv.Sdr/Asv.Sdr.csproj -c Release
dotnet pack src/Asv.Sdr.Simulate/Asv.Sdr.Simulate.csproj -c Release
dotnet pack src/Asv.Sdr.LimeSdr/Asv.Sdr.LimeSdr.csproj -c Release
dotnet pack src/Asv.Sdr.SignalHound/Asv.Sdr.SignalHound.csproj -c Release

On Windows, the repository also contains build.bat, which updates project versions from the latest Git tag, restores, builds, and packs the main packages.

Quick Start

Read the Getting Started guide for a complete hardware-free example. The FFT and Measurements guide shows streaming AM extraction with explicit stage disposal.

GetAm expects FFT coefficients. Use the processing order source.Sample(...) → Fft1d() → GetAm(sampleRate, 90, 150). Sampler block sizes count scalar values; each IQ pair occupies two values.

FFT Backend

Fft1d() uses ALGLIB by default to preserve existing behavior. You can switch the FFT implementation globally before creating FFT pipeline stages:

using Asv.Sdr;

ReaderIqFftSettings.Implementation = ReaderIqFftImplementation.Alglib;
// or
ReaderIqFftSettings.Implementation = ReaderIqFftImplementation.Managed;
// or
ReaderIqFftSettings.Implementation = ReaderIqFftImplementation.ManagedArmOptimized;

Available implementations:

• Alglib: existing ALGLIB-backed FFT implementation.
• Managed: built-in managed FFT implementation. It uses a radix-2 FFT for power-of-two transform sizes and a DFT fallback for other sizes.
• ManagedArmOptimized: managed FFT with an ARM64 AdvSimd fast path for radix-2 transforms. It falls back to the regular managed implementation when ARM64 SIMD is not available.

ADS-B Processing

Asv.Sdr includes helpers for building an ADS-B processing chain from IQ magnitude samples.

See ADS-B and Mode-S for parser setup, sample-rate constraints, and a complete chain attachment example. Magnitude() preserves interleaved (magnitude, 0) values; the pulse detector reads the even indices and emits a packed scalar pulse buffer. ProcessSample consumes 0/1 half-bit decisions including the preamble, rather than packed frame bytes.

Repository Layout

.
|-- build.bat
|-- publish_nuget.bat
|-- src
|   |-- Asv.Sdr
|   |-- Asv.Sdr.Simulate
|   |-- Asv.Sdr.LimeSdr
|   |-- Asv.Sdr.SignalHound
|   |-- Asv.Sdr.Msi
|   |-- Asv.Sdr.Evsg
|   |-- Asv.Sdr.Gspn1
|   |-- Asv.Sdr.AdSdr
|   |-- Asv.Sdr.Test
|   `-- Asv.Sdr.slnx
`-- LICENSE

Versioning

Package versions are defined in project files and shared build properties under src/Directory.Build.props. The documentation build reads ProductVersion from that file and displays it as the documentation version. Changes to the file trigger the documentation workflow.

The Windows build script uses the latest Git tag via git describe --tags and the dotnet-setversion tool before packing the main packages.

License

This project is licensed under the MIT License. See LICENSE for details.