GNSS-Aided Passive RF Emitter Localisation System

Project overview

Radio direction finding estimates the direction from which an electromagnetic signal arrives. A single direction-finding measurement provides a bearing rather than a unique location, meaning the transmitter may lie anywhere along the measured bearing line.

This project combines a directional antenna and software-defined radio receiver with GNSS positioning and antenna-heading measurement. Each observation records the receiver location, direction of peak received signal strength, GNSS quality information and a timestamp. Bearings collected from multiple positions are fused to estimate the location of a known and authorised test transmitter.

The system is designed as a passive, receive-only engineering demonstrator. It does not interfere with, decode private communications from or transmit toward any external system.

Engineering motivation

The engineering problem is not simply detecting a signal. It is producing a repeatable and defensible location estimate despite GNSS error, heading error, antenna imperfections, receiver noise, poor observation geometry and multipath propagation.

Objectives

1. Measure received signal strength while rotating a directional antenna.
2. Estimate the direction of peak signal arrival.
3. Record the GNSS position and timestamp of every observation.
4. Measure the physical heading of the antenna.
5. Combine at least three bearings into a transmitter-location estimate.
6. Quantify bearing and localisation error.
7. Detect low-confidence GNSS observations.
8. Present measurements on an interactive map.
9. Document calibration, design decisions and limitations.
10. Maintain receive-only operation throughout testing.

System requirements

Target status: All performance figures below are initial engineering targets. They are not achieved values and will be replaced or qualified after controlled testing.

System architecture

The architecture keeps raw acquisition separate from estimation. That separation supports replay, comparison of algorithms and evidence-based fault investigation.

Subsystems

Bearing estimation and processing

1. Configure SDR
2. Estimate signal power
3. Associate power with antenna angle
4. Smooth angular response
5. Estimate strongest valid peak
6. Attach GNSS and integrity data
7. Store raw observation
8. Fuse multiple bearings
9. Display solution and confidence

Illustrative angular response

The intended algorithm removes invalid samples, estimates a local noise floor, smooths the angular response, identifies the strongest valid peak, fits the neighbourhood of that peak, estimates a bearing and uncertainty, then compares clockwise and anticlockwise sweeps.

Uncertainty will be derived from repeatability, response-peak width, heading uncertainty and sensitivity to multipath—not from an arbitrary fixed margin.

Observation data model

timestamp,latitude,longitude,altitude,satellite_count,hdop,gnss_accuracy,fix_type,antenna_heading,heading_uncertainty,centre_frequency,rssi,noise_floor,snr,estimated_bearing,bearing_uncertainty,observation_status
2030-01-01T12:00:00Z,51.500000,-0.120000,35.2,12,0.9,4.8,3D,41.6,2.5,433920000,-52.1,-78.4,26.3,42.0,7.0,accepted

Localisation mathematics

For receiver position pᵢ = [xᵢ, yᵢ]ᵀ and measured clockwise bearing θᵢ, define the bearing direction and its perpendicular normal:

A candidate transmitter position x should lie close to each bearing line. Its perpendicular residual is:

The weighted least-squares solution minimises the total squared residual:

Greater weight can be assigned to observations with better SNR, narrower peaks, stronger GNSS quality, lower heading uncertainty and useful crossing geometry. Weighting cannot rescue fundamentally poor geometry; condition and residual diagnostics remain visible.

GNSS integrity monitoring

This work uses the term integrity monitoring: checks that identify questionable observations and adjust their influence. It does not claim complete GNSS resilience or assured navigation.

GNSS integrity score: 0.87
Bearing confidence: 0.76
Combined weight: 0.66
Observation status: accepted

Hardware and integration

The first portable assembly is expected to combine an SDR, directional antenna, GNSS receiver, heading interface, processing computer and power subsystem on a fixed rotating mount. Cable strain relief, RF connector integrity, power distribution, grounding, enclosure geometry and sensor placement are treated as engineering constraints rather than presentation details.

Experimental methodology and test plan

Open-field testing

Establish baseline bearing accuracy, minimise reflections and validate the core multi-position localisation method with an owned or authorised test transmitter.

Multipath testing

Repeat controlled measurements near buildings, fences, parked vehicles and other reflective structures to compare open-field and reflective-environment performance.

Ethical and legal test constraint: All trials remain passive and receive-only, use owned or explicitly authorised transmitters, follow applicable spectrum rules, and do not intercept private communications or locate individuals.

Results framework

This section is intentionally complete in structure but empty of experimental claims. Validated plots will be generated from versioned local JSON or CSV-derived data after controlled testing.

Engineering decisions

Failure and iteration log

These entries demonstrate the information architecture only. They are not claimed development events. Real version, evidence, root cause, change, outcome, date and media will replace them as the system develops.

Limitations, ethical scope and future work

Future improvements

Repository and project artefacts

The proposed structure separates hardware, firmware, software, raw evidence and processed results so changes remain traceable.