SDR-Shark

Spectrum collector and analysis module for the RTG Spectrum platform.

SDR-Shark provides the platform's SDR-facing spectrum workflow: synthetic or live IQ sessions, FFT, waterfall, scanner controls, decoded activity, and RF/ML hooks that can feed shared timeline and operator-review layers.

Open GitHub RF Signal Classification

A repeatable SDR collector workflow reviewers can run without radio hardware.

The public demo generates deterministic synthetic IQ data locally, starts the SDR-Shark backend and frontend, activates replay, and exposes the browser workflow used for spectrum visualization and signal review. In the RTG Spectrum architecture, SDR-Shark is the spectrum collector and analysis module that can publish evidence into broader normalization, correlation, and operator workflows.

Run the public Docker demo in one command:

docker compose -f docker-compose.demo.yml up --build

Deterministic synthetic replay, packaged for Docker.

The demo session uses synthetic 2.4 GHz carriers, moving noise floor, burst activity, and a Bluetooth-like frequency-hopping emitter that moves across a stationary receiver view. It is intentionally synthetic and public-safe; no captured RF environment data is tracked in the repository.

Run It

git clone https://github.com/rtg-engineering/sdr-shark.git
cd sdr-shark
docker compose -f docker-compose.demo.yml up --build

Open http://<machine-ip>:8080 on the local network. Stop it with docker compose -f docker-compose.demo.yml down.

Verify It

curl http://127.0.0.1:8080/api/iq/sessions
curl http://127.0.0.1:8080/api/iq/replay/status
curl http://127.0.0.1:8080/api/data

A healthy demo reports active replay and non-empty FFT and waterfall arrays from the data endpoint.

Browser-based spectrum, waterfall, and signal activity views.

SDR-Shark presents RF software as an operator-facing web application: spectrum traces, waterfall history, decoded activity cards, scanner configuration, settings dialogs, and analysis workflows are visible from the browser rather than hidden inside scripts or notebooks. Those views provide the SDR-facing evidence layer for broader RF signal classification, collect, normalize, correlate, baseline, and detect-change workflows.

Move SDR work from lab-only scripts into repeatable prototype evidence.

Reduce manual RF triage

Use replayable FFT, waterfall, activity cards, and scanner views to review signal behavior faster.

Prototype without hardware gates

Package deterministic demo data so stakeholders can evaluate the workflow before every radio path is available.

Feed shared analysis

Turn spectrum observations into event-ready evidence that can flow into Trace Analyzer and other review tools.

Fresh screenshots from the synthetic replay workflow.

These images were captured from the public Docker demo using the deterministic public-demo-2p4ghz session. The receiver stays centered while the synthetic Bluetooth-like emitter hops across the passband.

SDR-Shark public Docker demo full dashboard with synthetic RF replay — Full browser dashboard showing stationary receiver state, synthetic carriers, FFT, waterfall, and signal activity context.

SDR-Shark synthetic replay spectrum and waterfall detail — Synthetic 2.4 GHz replay detail with a Bluetooth-like hopping emitter, narrowband carriers, moving noise floor, and waterfall history.

SDR-Shark signal activity side panel from public replay demo — Signal activity panel wired to the same operator interface used for decoded protocol workflows.

Where this applies.

Collector Module

Prototype receive pipelines, scanner dwell plans, replay tools, dashboards, and integration glue for software-defined radio systems.

Event-Ready Evidence

Turn IQ streams, FFT products, detected activity, and metadata into reviewable interfaces and normalized evidence artifacts.

RF/ML Integration

Prepare hooks for classifier workflows, feature extraction, model-backed labeling, and signal-family exploration.

Demo Packaging

Create deterministic, public-safe demonstrations that let stakeholders evaluate a working system without lab hardware.