Investigating the Galileo Satellite Navigation System Outage with a LimeSDR

Galileo is a European Union owned satellite navigation system. Galileo was created so that the EU does not need to rely on the US GPS or the Russian GLONASS satellites, as there is no guarantee that these systems won't be purposely turned off or degraded by their governments at any time.

Unfortunately since July 11 the Galileo system has been out of service. Not much information about the outage has been provided, but it appears to be related to problems with the Italian ground based Precise Timing Facility which consists of two ultra high precision atomic clocks that keep the Galileo systems' reference time. (We note that recently within the last few hours of this post, most satellites seem to have come back into operational status, but the EGSA website still reports an outage.)

Over on his blog, Daniel Estevez has been using his LimeSDR and a small patch antenna to gather some more information about the outage directly from the Galileo satellites. His investigations found that the modulation and signal itself are still working correctly. However, by using the GNSS-SDR software to investigate the signal data he was able to obtain the ephemeris, and see that the ephemeris is stuck in the past. The ephemeris data is used to calculate compensations for orbital drift and without frequent ephermis updates, orbital errors add up within hours resulting in poor positioning accuracy. In order to generate the ephermis, the Precise Timing Facility must be operational.

Daniel's post goes into further technical details about the information he's collected, and it's definitely an interesting read. One interesting bit of information that you can read from his post explains why the service has gone from initially just heavily degraded accuracy from July 11, to completely nonsense results from July 15 onwards.

Upcoming KerberosSDR Software Updates: Automatically Estimate TX Location and Navigate There

KerberosSDR is our 4x Coherent RTL-SDR that we've developed together with Othernet. It can be used for tasks such as direction finding and passive radar. KerberosSDR was successfully crowdfunded over on Indiegogo, and the first batch has already been shipped. Currently we are taking discounted pre-orders for a second production batch on Indiegogo. Please note that the discounted pricing will expire when we ship, which according to the manufacturing schedule should be next month, so please get in quick if you're interested!

If you'd like to back the KerberosSDR project and purchase a unit, please see our Indiegogo page.

Below are some recent updates to the project:

Android App Software Improvements

The Android App allows a KerberosSDR user to drive around in a car, collecting angle of arrival data for a signal. Driving around and collecting multiple data points solves the multipath issue. In a single location it is possible for a signal's direction of arrival to be skewed or incorrect as it can bounce off multiple surfaces and appear to be arriving from a wrong direction. If we collect data from many locations, we can average out the multipath.

We've recently been working on improvements to the direction finding capabilities of the KerberosSDR, and in particular to our free Android App which records and plots data from the KerberosSDR server. We are still testing and finalizing these new features, but hope to release the updated app before the end of this month.

Recently added features to the app include:

  • Added the ability to determine the estimated location of a transmitter, providing there has been sufficient data collected.
  • Added a heatmap grid of the collected data which can be used to determine where most angle lines cross. Can take into account RF power data too.
  • Added the ability for the software to automatically navigate you to the estimated TX location via MapBox GPS turn by turn navigation.

Bellow are screenshots showing some of the new features. In this experiment we located an 858 MHz TETRA transmit tower. Initially the app will navigate you to the edge of the grid, in the direction that most DoA lines are pointing to. When there is sufficient data to be able to confidently pinpoint the TX location, it will begin navigating you to the estimated location. In the screenshots the placemarker represents the known location of the transmitter, and the circles indicate the location estimated from direction finding.

Below is screenshots from a 415 MHz DMR tower that we located with KerberosSDR. The antenna array was purposely kept small, with a diameter of only 12cm. Even with the small antenna array we were able to pinpoint the transmitter down to about 100 - 200 meters.

The app should also now be able to handle intermittent signals, via a squelch filtering function, although this has not been fully tested yet.

In order to navigate you must have a 3G/4G data plan on your phone, and your phone must have the ability to create a WiFi hotspot. The KerberosSDR server running on a Pi 3 or similar will then automatically connect to a WiFi hotspot named "KerberosSDR" running on your phone and provide data to the app via WiFi.

Batch 2 Manufacturing Updates

Batch 2 production is in full swing, and at the moment we're expecting completion by mid August. This batch will ship directly from China, so we should be able to ship them off fairly quickly rather than needing to first wait for them to arrive in the USA.

Magnetic Whip Antennas

We have been disappointed that it has been difficult to find low cost but good quality magnetic whip antennas to use with KerberosSDR and vehicles. The quality of antennas used in direction finding equipment can matter, as any signals leaking into the coax, or radiation pattern skew can affect results. We are working on sourcing some high quality magnetic whip antennas that have good ground coupling. These will be sold at a reasonable price on our store.

Future Updates

We are still working on improving the server software further too and future updates will include things like the ability to notch out unwanted signals during phase calibration, a simplified DoA set up wizard, an improved buffering scheme so that additional data and processing gain can be applied, and more.

The Raspberry Pi 4 looks to be an excellent candidate to be used with the KerberosSDR. We will begin releasing ready to use images for the Pi 4 in the future.

Thanks!

Every sale of a KerberosSDR helps fund further developments to the software and possible future iterations of the hardware. So we'd like to thank all backers once again!

Signals and Bits A New SDR Podcast by the President of the GNU Radio Project

[Ben Hilburn] the president of the [GNU Radio Project] has recently started a new podcast called [Signals and Bits]. If you were unaware, GNU Radio is the defacto open source framework for implementing digital signal processing code. Without it, many SDR programs that we take for granted may have never been developed as it is responsible for a lot of community DSP knowledge and algorithm development.

This podcast is scheduled for a new release every Wednesday and will be composed in an interview style focusing on a multitude of topics from Software Defined Radio to Spectrum Enforcement, Radio Astronomy and so much more.

In the first episode Ben interviews Harold Giddings AKA Corrosive of [Signals Everywhere] where they discuss the state of Software-Defined radio and how he got started with radio communications having come from an IT/Computer Networking background.

Ben has already pre-recorded several episodes which will ensure great content is always just around the corner. Ben would love it if you could also send feedback his way over on the [Signals and Bits Twitter] page.

Ben Hilburn President of GNU Radio (Left), Harold Giddings AKA Corrosive of Signals Everywhere (Right)

Artemis 3 Released: Offline Signal Identification Database

The Signal ID Wiki (sigidwiki) is our sister site that we started a few years ago as a way to collect and catalog various types of signals that an SDR user might see and hear on the airwaves. The idea is that a user could search the database to learn about and identify unknown signals. Over time it has grown significantly, with now almost 400 known signals with both waterfall images and sound samples available in the database. Special thanks to lead admin Carl Colena for maintaining and playing a huge role in the databases' growth.

Artemis is an open source Windows/Linux/MacOS compatible application initially programmed by Marco Dalla Tiezza. It brings the sigidwiki website into an offline searchable database with an easy to use UI. Today version 3.0 was released to the public. The new version has been completely rewritten from scratch in Python, as the previous versions were written in BASIC (a now abandoned programming language). The new version has an improved UI, and paves the path for future improvements. 

Marco notes that in the future they hope to add an Autocorrelation function, which might help users automatically identify certain types of repetitive signals simply by playing the raw audio into Artemis.

Note that in order to download the software you will need to sign up to their forum, which is free.

Artemis 3.0 Screenshot
Artemis 3.0 Screenshot

rtlSpectrum: A New GUI for rtl_power

Thank you to Andrey for submitting news about the release of his new open source program called "rtlSpectrum". rtlSpectrum is a GUI for rtl_power, which allows you to scan a wide swath of bandwidth with an RTL-SDR dongle. The scan can then be plotted with rtlSpectrum, and peaks of activity can then be determined. Some of the features include:

  • load from .csv file produced by rtl_power
  • run rtl_power directly. it should be available in the $PATH
  • add multiple graphs for analysis
  • subtract one graph from another
  • save/export graph in the rtl_power based format

In particular, the ability to subtract one graph from another is useful for analyzing filters. Andrey has posted about doing this with rtlSpectrum, a noise source and an RTL-SDR V3 over on his blog (note that the blog is in Russian, so please use Google Translate if necessary). He notes how the dynamic range of the RTL-SDR is limited, so that the true blocking power of a filter cannot be determined, but it is enough to see the shape and frequency response.

rtlSpectrum Screenshots
rtlSpectrum Screenshots

SignalsEverywhere: Creating a 2.4 GHz Amateur Television Station with a Barbie Camera

In this episode Corrosive from the SignalsEverywhere YouTube channel begins a series on reverse engineering an old 90's Barbie branded 2.4 GHz wireless camcorder toy, and using it for Amateur Radio TV. The camcorder toy consists of the wireless camera, and a base station that plugs into a TV.

After taking the camcorder apart Corrosive discovered a potentiometer on the PCB which allows the transmit frequency to be adjusted, and that the camera's CCD sensor can actually output composite video, possibly allowing for improved video quality. In addition he found the datasheet for the main FM demodulator chip on board the base station, and saw that it is designed to operate from 350 to 550 MHz. So he speculates that elsewhere in the circuit is a 2.4 GHz downconverter which may be useful for other projects too.

Two Antenna Diversity For the SDRPlay RSPDuo now Available in SDRUno

The SDRplay RSPDuo is a 14-bit dual tuner software defined radio capable of tuning between 1 kHz - 2 GHz. It's defining feature is that it has two receivers in one radio, which should allow for interesting phase coherent applications such as diversity.

In the latest v1.32 release of SDRUno a diversity feature has been added. Diversity reception was demo'd back in May at Hamvention, and we have a previous post with video about that. Diversity works by subtracting or adding two signals from the two receivers running independent antennas. The result is an up to 3 dB increase in SNR, and better performance with fading signals. They write:

From V1.32 onwards, MRC (Maximal Ratio Combining) Diversity is supported using the RSPduo. MRC Diversity can be used to combine the 2 tuner input streams together to potentially improved the SNR (signal to noise ratio). The same frequency is used for both tuners in the RSPduo and the gain can be adjusted either on each tuner independently or locked together (the default method).

Diversity mode is enabled by clicking on the RSPduo MODE dropdown and select DIVERSITY. Make sure both the 50 ohm ports are connected to the correct input source and note that the HiZ port is not available for Diversity mode. Trying to use the HiZ port will result in an error message being displayed.

Diversity Controls in SDRUno
Diversity Controls in SDRUno

Meteor M N2-2 Successfully Launched: Awaiting LRPT Signal

On July 5 the Russian Meteor M N2-2 weather satellite was successfully launched into orbit and appears to be healthy. The LRPT weather camera signal is not yet broadcasting however, and we expect it to still take roughly 1-2 months before it begins (if all goes well) as satellites typically run through a long list of qualification tests before becoming operational. During this time there may be broadcasts of test patterns that can be caught. Meteor M N2-2 can currently be tracked in Orbitron and online at N2YO.

To try and dispel any confusion over the naming scheme, "Meteor M N2" is the currently operational LRPT satellite. "Meteor M N2-1" unfortunately failed in 2017 as it did not separate from the rocket. "Meteor M N2-2" is the new satellite which has just been successfully launched. Meteor M N2 and M N2-2 is often abbreviated as just "Meteor M2" and "Meteor M2-2". In the past there was Meteor M N1, but this satellite is no longer operational. We have upcoming launches for Meteor M2-3, M2-4, MP-1 and M3 to look forward to which are scheduled for 2020 and 2021.

Back on June 28 we posted about how Meteor M2 was experiencing orientation issues for a few days. Those issues appear to have been rectified now. Hopefully if M2 remains stable we'll have two Meteor LRPT weather satellites to receive alongside the three NOAA APT satellites.

If you're interested, there were also several other payloads onboard the rocket carrying M2-2, including a low cost Czech experimental cubesat called Lucky7 whose telemetry can be received in the amateur radio band at 437.525 MHz. There is an onboard camera too, but no details on how to receive it yet.

Soyuz Rocket carrying Meteor M2-2 Launched successfully. [Source: spaceflightnow]