css body, p, .post-body { font-family: 'Google Sans Text', sans-serif !important; } /* Apply Google Sans to Post Titles and Headings */ h1, h2, h3, h4, h5, h6, .post-title, .post h2 { font-family: 'Google Sans', sans-serif; font-weight: 500; }
Showing posts with label AI. Show all posts
Showing posts with label AI. Show all posts

Wednesday, July 15, 2026

South Africa's own Oscar- DMR 1 Satellite. Will this DMR Satellite ever go to Space?


Image:  AI (Click on image for larger view.) 

I wrote several articles in the past where I refer to innovation in Amateur Radio.  This morning a thought came to mind on why is there currently not a satellite with a DMR Transponder (repeater) up in space?  I was trying to think "out of the box" and look at ways and means to turn my thoughts into reality, if there is such a possibility.

Thinking "out of the box" is absolutely a great way to approach innovation in amateur radio.  Amateur Radio has a rich history of development driven entirely by amateurs experimenting with limited resources and unique constraints. 

However, true innovation in this hobby relies on a balance between unconventional thinking and foundational science. 

Why Out-of-the-Box Thinking Works

  • Resourcefulness: Limited power regulations and frequency bands force you to find clever ways to maximize efficiency.

  • Historical Precedent: Amateurs invented weak-signal digital modes (like FT8) and bounce signals off the moon (EME) because someone asked "what if?"

  • Cross-Pollination: Bringing concepts from computing, machine learning, or material science into radio often yields breakthrough results.

The Innovation Blueprint in Amateur Radio

To make your creative ideas successful, pair your out-of-the-box thinking with these structured approaches:

  • Master the Fundamentals: You must understand Maxwell's equations, wave propagation, and circuit design to break the rules effectively.

  • Identify Real Constraints: True innovation solves a specific problem, such as reducing noise, shrinking antenna size, or bridging communication gaps during disasters.

  • Iterate and Test: Build prototypes, collect data, and use antenna analyzers or software simulation tools to prove your theories.

  • Share with the Community: Amateur radio thrives on open-source collaboration. Presenting your ideas on forums, Git repositories, or at club meetings helps refine them.

Prominent Areas Needing Innovation

If you are looking for modern challenges to apply your creativity, focus on these emerging sectors:

  • AI and Machine Learning: Using neural networks for predictive propagation filtering or automated signal decoding in dense noise.

  • Alternative Materials: Experimenting with meta materials, fractal geometry, or everyday conductive liquids for stealth antenna designs.

  • Digital Signal Processing (DSP): Developing open-source algorithms to extract ultra-weak signals from high-noise urban environments.

  • Green Energy Integration: Creating ultra-efficient, off-grid solar or kinetic power management systems for remote field operations.

Now reading the above my thoughts ran away with me regarding a DMR transponder (repeater / beacon / node / hotspot) payload on a South African satellite.  Can it be done and is it possible to communicate using  DMR via a satellite?  Is there a satellite specifically named "Oscar-DMR 1" in existence or being built in South Africa's space programme or amateur radio history. Currently this type of communications is not possible.

If you are waiting for a homegrown South African satellite or an OSCAR (Orbiting Satellite Carrying Amateur Radio) series spacecraft to launch, here is the factual reality of what has actually happened and what is currently in progress.

The Real South African Satellites

South Africa has already successfully launched multiple satellites. If you are looking for local aerospace achievements "seeing the daylight," they have already made it to space: 

  • SUNSAT (SO-35): Launched in 1999, SUNSAT was South Africa's very first satellite, built by Stellenbosch University. Critically, it carried an amateur radio payload and was officially designated as OSCAR 35 (SO-35) by AMSAT. It saw plenty of daylight before its mission ended. 

  • ZACUBE-1 & ZACUBE-2: Developed by the Cape Peninsula University of Technology (CPUT). ZACUBE-1 (TshepisoSAT) launched in 2013, and ZACUBE-2 launched in 2018. 

  • MDASat-1 Constellation: In January 2022, South Africa successfully launched a three-nanosatellite constellation via a SpaceX Falcon 9 rocket. These operational maritime domain awareness satellites track shipping traffic off the South African coast. 

ZS1I created the fictional "DMR 1" Satellite Name 

The term DMR stands for Digital Mobile Radio, which is a widely popular land-based digital protocol used by radio enthusiasts and businesses across South Africa. 

  • Terrestrial, Not Space: Amateur radio operators in SA heavily utilize the South Africa DMR Repeater Network. This network relies on ground-based repeater towers, not dedicated South African "DMR satellites." 

  • Conflation with Commercial Satellites: You may be thinking of commercial mobile satellite services or push-to-talk satellite radios that interface with DMR-style dispatch systems on the ground. 

Future South African Space Missions

If you are wondering about the next major government-backed leap into orbit, the Department of Science and Innovation has active plans: 

  • National Communication Satellite: The government has been exploring multi-billion-rand plans to acquire or launch a dedicated communications satellite to bridge the digital divide and reduce reliance on international space entities.  However this look like a very "far in the future" project with many logistical and financial issues to first solve.

  • Deep Space Ground Tracking: While not a satellite itself, South Africa broke ground on a massive, state-of-the-art Deep-Space Ground Station in Matjiesfontein (Karoo), built in partnership with NASA to track future lunar missions. 

Sadly you cannot work DMR (Digital Mobile Radio) directly through orbiting amateur radio satellites.  Hopefully by means of innovation in technology my thought on building such is satellite is not far fetched and will not forever just be a thought.  In South Africa building such a satellite will be problematic but that is a topic for another time.

Let's see why this can or cannot currently be implemented.  

You can absolutely build a physical DMR repeater and launch it into orbit on a satellite. However, doing so introduces a severe physics and timing challenge that standard DMR protocols are not designed to handle. 

The core issue is the speed of light and the Doppler effect.

The Timing Problem (TDMA Breakdown)

DMR relies on TDMA (Time Division Multiple Access). It divides a single frequency channel into two distinct time slots (Slot 1 and Slot 2). 

  • The Rule: Each radio must transmit in an incredibly precise window—exactly 30 milliseconds long.

  • The Margin: The standard protocol only accounts for standard terrestrial distances, leaving a guard timing buffer of about 1.25 milliseconds to handle propagation delay. 

When a satellite is orbiting overhead (even a Low Earth Orbit, or LEO satellite at ~500 km), the distance from the ground station to the spacecraft changes continuously and rapidly. Because the radio waves must travel hundreds of kilometers to space and back, the propagation delay exceeds that 1.25 ms guard window.

Consequently, your radio's packet arrives late, shifts out of its slot, and bleeds into the adjacent time slot. This completely breaks the synchronization, causing the repeater's onboard computer to reject the handshake. 

The Frequency Problem (Doppler Shift)

DMR uses 4FSK digital modulation. It relies on precisely mapping four distinct, narrow frequency shifts to represent binary data. 

Because a LEO satellite travels at roughly 27,000 km/h, the frequency shifts dramatically as it approaches and moves away from you. This Doppler shift warps the digital signal. While a human ear can decode an uncorrected analog FM signal through a bit of static, a digital DMR modem will see the warped 4FSK signal as corrupted gibberish and refuse to decode it. 

How to Make a Space-DMR Repeater Work

If an amateur radio group or space agency wanted to make a true DMR satellite work, they would have to implement one of two workarounds:

  1. Modify the Radio Firmware (Software Solution)
    The ground station's DMR radio would need custom firmware capable of predicting the satellite's exact orbit. The radio would then have to continuously alter its timing (transmit early or late to hit the slot perfectly) and automatically adjust its frequency to cancel out the Doppler shift in real-time.
     

  2. Put the Spacecraft in a Geostationary Orbit (Hardware Solution)
    If you put the DMR repeater on a geostationary satellite (35,786 km above Earth), the satellite remains stationary relative to the ground. This eliminates the Doppler shift entirely. While the time delay would be much larger, it would be
    constant, allowing engineers to build custom terrestrial radios with a massive, fixed timing buffer specifically for space.
     

Consolation Prize 

There is currently a "consolation prize" on how you can use your DMR Radio to connect to satellites indirectly.

You can use your DMR radio to connect to satellites indirectly by talking through an MMDVM hotspot (or a local digital repeater) connected to the internet. From there, your signal is routed to space through a commercial geostationary satellite (such as QO-100) using an up/down converter, a dish, and an SDR (Software Defined Radio). 

Unlikely that a dedicated amateur satellite named "Oscar - DMR 1" will be built in South Africa

It is highly unlikely that a dedicated amateur satellite named "Oscar - DMR 1" will be built specifically for standard DMR voice communications in South Africa. While amateur radio organizations like AMSAT constantly develop new spacecraft, standard commercial DMR protocol is fundamentally incompatible with the physics of Low Earth Orbit (LEO) satellites.

The Geostationary Exception, there is hope!! 

The only way a true DMR transponder could work in space is on a Geostationary (GEO) satellite like QO-100. Because GEO satellites remain completely stationary relative to the Earth's surface, there is zero Doppler shift or changing propagation delay. While there is no official "Oscar - DMR 1" payload planned, experimental digital voice links are routinely tested via GEO transponders using specialized ground stations.  More on this in a future article once I put on my "out of the box" and "innovation" hat.

Was this article a waste of time?  NO definitely not.  I now have more questions than answers that I will be looking into.

ED. This article was compiled by ZS1I with the assistance of AI. 

Saturday, June 27, 2026

Amateur Radio in the age of AI

Video:  Dr. Paris Buttfield-Addison (VK7SYN) discusses "Artificial Intelligence & Machine Learning in Amateur Radio," what AI actually is & demonstrates the potential for AI to enhance amateur radio.  -  Ham Radio DX 
Amateur Radio in the Age of AI 
Artificial intelligence is revolutionizing amateur radio by automating routine tasks, enhancing signal processing, and optimizing contest strategies. Far from rendering the hobby obsolete, AI acts as a powerful operating assistant—improving noise filtering, expanding accessibility for operators with disabilities, and advancing global spectrum experimentation. 
Key Applications of AI in Ham Radio
  • Signal Processing & Noise Reduction: AI algorithms are increasingly integrated into software-defined radios (SDRs) and digital signal processors (DSP). They can intelligently filter out background noise, isolate weak signals in harsh atmospheric conditions, and enhance audio clarity. 
  • Contest Strategy & Logging: AI analyzes massive datasets from the DX Cluster to provide real-time recommendations on rare stations, predict optimal band frequencies, and optimize your overall score during major contesting events. 
  • Accessibility & Voice Control: Machine learning models assist operators with speech impairments or visual limitations to participate in digital modes (like FT8) through automated text-to-speech, voice control, and digitized voice generation. 
  • Propagation Forecasting: AI systems process historical and real-time space weather, solar flux index (SFI), and geomagnetic data to generate highly accurate HF (High Frequency) propagation predictions.
What AI Cannot Replace
While AI can help you hunt down contacts or log QSOs, the core of amateur radio remains human. The technology cannot replicate the thrill of building physical antennas, improvising off-grid communications during emergencies, or the tactile feel of tuning a radio. The regulatory framework for amateur licensing and transmitting—managed globally by bodies like the ITU—still requires a licensed human operator at the helm. 
Now lets look a little deeper into this sometimes controversial topic. 
The application of artificial intelligence and machine learning in amateur radio has transitioned from conceptual experimentation into real-world software tool-chains and radio hardware. AI operates as a powerful algorithmic layer that interfaces with the physical environment, processing massive amounts of telemetry data and raw RF (Radio Frequency) audio streams. 
The primary technical areas where AI is creating the most significant impact include advanced digital signal processing, dynamic ionospheric modeling, and cognitive station automation. 

1. Neural Networks & Advanced Digital Signal Processing (DSP)
Traditional DSP relies on hard-coded mathematical rules (like fixed Bandpass or Notch filters) to clean up signals. AI replaces or augments this with recurrent neural networks (RNNs) and adaptive filters that train on millions of noisy audio samples. 
  • Intelligent Noise Isolation: AI filters can dynamically distinguish between human voice, Morse code (CW), and ambient localized interference—such as EMI from solar panel inverters, power grids, or switching power supplies. It subtracts the noise in real time, making borderline unreadable signals intelligible. 
  • Automatic Signal Classification: Using low-power hardware (such as a Raspberry Pi paired with an RTL-SDR dongle), AI algorithms use open-source pipelines to instantly identify, classify, and isolate specific modulation types (e.g., APRS, FT8, FM, or satellite beacons) across wide swaths of the radio spectrum. 
2. Predictive Propagation and "Big Data" Ionospheric Modeling
Predicting whether an HF (High Frequency) signal will bounce off the ionosphere to reach a specific continent has historically relied on static monthly median models like VOACAP. AI shifts this to real-time, fluid forecasting: 
  • Telemetry Integration: Machine learning algorithms continuously ingest live data streams, including Solar Flux Index (SFI), geomagnetic activity (K-index, A-index), coronal mass ejection alerts, and planetary ionosonde readouts. 
  • Crowdsourced Spot Mapping: Modern AI architectures collect hundreds of thousands of daily data points from networks like the Reverse Beacon Network (RBN) and DX clusters. By analyzing the paths where signals are actually getting through right now, the AI builds deep-learning models to map out precise, real-time RF "micro-openings" on the bands. 
3. Smart Contesting, Automated Logging, and Strategy
During radio contesting—where the goal is to make as many rapid-fire contacts as possible—AI functions as a digital co-pilot. 
  • Predictive Spotting & Hunting: AI systems analyze cluster feeds to prioritize rare DX stations based on your station's historical capabilities, antenna trajectory, and local terrain limitations. It advises when to switch bands or call a specific frequency before the band opening disappears. 
  • Automated Call Translation: In weak-signal scenarios or heavy pileups, AI assists in audio decoding. Generative audio tools can infill missing packets of voice transmissions, predicting a call sign's broken suffix or prefix based on global license databases and phonetic speech patterns. 
4. Accessibility and Cognitive Radio Control
AI lowers the physical barriers to entry for disabled, aging, or speech-impaired operators, ensuring inclusivity in the amateur community. 
  • Speech and Language Translation: Real-time translation models allow operators of different nationalities to converse smoothly via voice. For operators with localized speech impairments, AI can map inconsistent vocal inputs into synthesized, digitized voices that cleanly trigger SSB (Single Side-band) transmitters. 
  • No-Code CW Assistants: Machine learning toolsets are being developed to interpret high-speed, poorly spaced, or drifting manual Morse code ("fists"). This translates raw audio into readable text on a screen without requiring the operator to master the code by ear. 

Comparison: Traditional vs. AI-Enhanced Radio Operation
Feature Traditional Amateur RadioAI-Enhanced Amateur Radio
Noise FilteringManual adjustments of RF gain, notch filters, and fixed audio DSP width.Dynamic neural networks that isolate human voice or code from background electrical hums.
Band HuntingManual tuning across a VFO dial or tracking simple text-based DX cluster alerts.Predictive spectrum scanning prioritizing frequencies based on real-time solar telemetry.
Digital DecodingExact mathematical pattern-matching; fails if signal drops below the hard theoretical noise floor.Generative packet-filling and probabilistic decoding of compromised data streams.
Shack MaintenanceManual reading of complex paper schematics to build antennas or debug circuitry.Computer vision and LLMs that troubleshoot physical circuit designs or guide antenna cuts via photo inputs.

From the beginning, amateur radio has connected people with reliable information and companionship, including in the most difficult moments during emergencies or disasters.

In this new era, AI must remain a tool to serve that mission: helping radio amateurs to assist more people, in more languages; never replacing the editorial responsibility for which communities rely on amateur radio stations during disasters.

World Radio Day, celebrated yearly on 13 February, honours the medium’s unique power to inform, connect and accompany people everywhere. 
The latest annual theme reminds us:  
AI is a tool, not a voice.”
We need to continue to preserve the Amateur Radio bands / airwaves as a valuable resource that enables this unique medium to thrive.

Ultimately, radio’s future depends on using AI to reaffirm and strengthen the human values that define the medium.
 

ED.  There is quite a few authors that contributed to this topic:

1. Dr. Paris Buttfield-Addison VK7SYN

2. Hayden P Honeywood VK7HH

3. Mario Maniewicz, Director, ITU Radiocommunication Bureau

4. Johan ZS1I

5.  AI

I would like to thank them for their input and outlook on AI.  AI was used as a tool, not a voice in this topic!   -  ZS1I 

Is Social Media killing amateur radio on-the-air activity?

  In a recent discussion with a fellow radio amateur we discussed the use of WhatsApp, Telegram and even Facebook as a means of forwarding ...