3 Understanding Node Data
The goal of this page is to explain the type of data being collected on your node and how it correlates to real objects in the sky.
Key Views Discussed:
- Radar and 10s Max Hold
- ADS-B
- Tracker Preview
You can access all of these views from your node at the URL: http://owl.local/.
What is bistatic radar? Hopefully you've read some of the basics on our website. This diagram from our site shows a basic picture of what's happening:
To detect an object, your reference antenna picks up the baseline pilot signal directly from the transmitter, and the system compares it against the surveillance antenna, which picks up reflections of that same signal off nearby objects. A moving object shifts the frequency of its reflection slightly due to the Doppler Effect. By cross-correlating the two channels, we can measure two things: the object's range (from the time delay between the direct signal and the reflection) and its relative velocity (from the Doppler frequency shift).
Doppler shift example:
The output of those measurements is shown in the Passive Radar live data plot.
The Passive Radar live data output plot is an instantaneous snapshot of the reflections that stand out with a strong signal-to-noise ratio relative to the baseline pilot signal. Rather than plotting absolute frequency, we plot the Doppler shift in Hz relative to the baseline signal. We also calculate the extra time it takes the reflected signal to travel from the object to the surveillance antenna, which gives us range in kilometers.
So if you're using a 200 MHz signal, the Doppler shift is measured relative to that 200 MHz carrier, up to about ±300 Hz. In other words, a single instantaneous detection might show a Doppler shift of +84 Hz (i.e. 200,000,084 Hz) and a range of 10 km, which is then plotted on the graph.
To take the visualization a step further, we don't just want a single instant in time — we want a 10s window of detections so we can see the flight path of an object. That's what a 10s max hold plot gives you a picture of.
Imagine a car driving past you. As it approaches, you hear its horn at a higher pitch because the sound waves are compressed — this puts the car in the 0 to +300 Hz upper half of the Doppler map. At its closest point of approach (CPA), the track crosses the 0 Hz line. As it drives away, the pitch drops because the waves are stretched, which puts it in the 0 to -300 Hz lower half of the map.
One limitation of a bistatic passive radar system is that it doesn't tell you the object's location on a map — only its range and speed. So when you see a detection, the object could be anywhere along a circle of that range around your node location.
In reality your antenna is probably directional, with a certain cone of detection. With some math we run in the node we are still able to derive which direction your antenna is probably facing, so you don't have to worry about the full circle around your location. There's still the question of where within that narrower arc the object actually is.
There are a few ways we work around this:
- ADS-B
- Triangulation
- Fleet Data
Automatic Dependent Surveillance–Broadcast (ADS-B) is a continuously updated data feed that gives you live information on cooperative, transponder-equipped aircraft in your sky. We use ADS-B as a source of truth to compare the objects we detect against the aircraft currently passing overhead.
A better-tuned system will be more accurate and show closer overlap with ADS-B for correlation:
We can then subtract out the known aircraft to judge whether a detection is more likely to be anomalous. Note that ADS-B is not completely comprehensive — some military flights won't appear.
Another way to pin down an object's true location relative to the observing node is to position two nodes close enough in pointing direction that both can detect the same object. These could both be on your property, or one could be on a neighbor's property a couple of miles away.
If they both detect an object, there are two theoretical points where it could be.
We can then do some math to interpolate which of the two points the object is most likely at.
We also plan to support triangulation using a single surveillance antenna picking up reflections from multiple towers. This isn't available yet, but keep an eye out for it in the future!
The final view we currently build at your node is the tracker view. It's still a work in progress, but the goal of this view is to highlight any objects that have flown overhead in the last 4 hours.
- Detections in Delay: Detection ranges over time
- Detections in Doppler: Detection Dopplers over time
- Detections in Delay-Doppler: Range vs. velocity scatter plot
- Spectrum: Reference channel spectrum currently being seen
- Timing: Processing time for each stage
Check out this page to see what things like birds, meteors, or UAP may look like on the RETINA system data.