Packet Radio

When looking at the historical context, “packet” in amateur radio is not just a technical unit, but a genuine paradigm shift. Starting in the late 1970s, with the birth of packet radio, amateur radio operators suddenly gained access to a remarkably advanced tool for the time: a form of packet-based digital communication that anticipated by years-if not decades-the concept of wireless data networking as we understand it today.

At a time when networks were still mostly wired and confined to academic or military environments, the ability to transmit structured data over radio, across an unstable and shared medium like the ether, represented something radically new. Packet radio can therefore be seen, in every sense, as the first practical form of wireless internetworking, built not by large infrastructures, but by a distributed community of experimenters.

It is precisely in this context that the concept of a “packet” as understood by radio amateurs was born. It is not simply a block of data, but a unit designed to survive a hostile environment: interference, noise, fading, and signal collisions. Unlike a continuous connection, packet introduces a fundamental idea: every piece of information is split into autonomous parts, each of which can travel independently, be verified, and, if necessary, retransmitted.

In the world of packet radio, this concept is implemented through the AX.25 protocol. But beyond formal details, what matters is the philosophy: each packet is a self-contained “messenger” carrying everything needed to reach its destination.

What a “packet” looks like

Without going into binary formalism, an AX.25 packet has a well-defined structure designed to operate over a real radio channel.

At the beginning and end there are flags, special sequences that allow the receiver to synchronize and identify where the packet starts and ends. Immediately after comes the address field: it contains first the destination callsign and then the sender’s callsign-counterintuitively in the opposite order one might expect-possibly followed by a list of digipeaters the packet must traverse. This structure, inherited from HDLC, allows intermediate nodes to immediately read the destination without parsing the entire frame.

Next comes the control field, which indicates the frame type (information, acknowledgment, connection request, etc.), and in information frames also a Protocol ID field, which specifies which higher-level protocol is encapsulated (TCP/IP, NET/ROM, and so on). Then comes the payload, the actual content: text, commands, network data.

Finally, a crucial element: the error-checking field, called the FCS (Frame Check Sequence), a 16-bit number computed over the entire frame that allows the receiver to verify whether the packet arrived intact. If it did not, it is discarded without ceremony. This seemingly simple mechanism is what makes reliable communication possible over an inherently unreliable medium like radio.

But how does all of this physically travel through the air? An AX.25 packet is, at its core, a sequence of bits-zeros and ones. To transmit them over radio, those bits must be converted into an audio signal, and the audio must modulate the radio.

The method used in classic packet radio is AFSK (Audio Frequency Shift Keying), the same principle used in RTTY, applied to packet. In practice, a zero and a one are represented by two distinct audio tones-on VHF the standard values are 1200 Hz for mark (bit 1) and 2200 Hz for space (bit 0), according to the Bell 202 standard-and these tones are transmitted via FM radio at 1200 baud. The speed is modest, comparable to an old 1980s telephone modem, but more than sufficient for the text and short data typical of packet traffic.

On HF, however, where bandwidth is narrower and propagation conditions more challenging, a much tighter shift is used: the most common tones are 1600 Hz for space and 1800 Hz for mark, separated by only 200 Hz, at a rate of 300 baud, transmitted in SSB. It is a simple and elegant system: the radio knows nothing about packets or protocols; it simply transmits audio tones, and everything else is handled by the TNC.

The role of the TNC

To make all this practical, a key component is required: the Terminal Node Controller, or TNC. It is essentially an “intelligent modem” that handles encoding and decoding of AX.25 packets, manages transmission timing, checks errors, and acts as a bridge between the computer and the radio.

In classic systems, the TNC was a dedicated hardware device, with its own microprocessor and firmware. On one side it connected to the computer via an RS-232 serial port; on the other, to the radio via audio and a PTT line. It handled all the “network” logic, while the computer was only responsible for the user interface-sometimes just a simple terminal.

Over time, this function has been progressively absorbed by software. Today, software TNCs are common, implemented in programs that use the PC’s sound card: modulation and demodulation are handled in audio, and the AX.25 protocol is entirely software-based. This has made packet far more accessible, eliminating the need for dedicated hardware and enabling implementations on virtually any PC, Raspberry Pi, or smartphone.

Packet radio as a network

One of the most innovative aspects of packet is that it is not just a transmission mode, but a full network system. Packets can be received and retransmitted by digipeaters, routed through multiple nodes, and temporarily stored in store-and-forward systems. This means there is no need for a direct connection between two stations: a message can traverse the network node by node, exactly like the Internet. And with systems such as Net/ROM and FlexNet, nodes could autonomously learn which paths were more reliable, introducing a surprisingly sophisticated form of dynamic routing for its time.

Evolution and current status

With the arrival of the 1990s and the spread of commercial Internet-much faster and easier to use-traditional packet radio gradually lost operational relevance. BBS systems, Net/ROM nodes, and distributed networks slowly disappeared or remained active only in niche contexts.

Today, packet has not entirely vanished, but survives mainly in an evolved and radically simplified form: APRS, the Automatic Packet Reporting System. The idea was developed in the early 1980s by Bob Bruninga, WB4APR, a researcher at the United States Naval Academy, who first used it in 1984 to track the position of horses in a 100-mile endurance race. The project was refined over the following years, officially named APRS in 1992, and became globally widespread with the arrival of consumer GPS.

APRS still uses AX.25, but with a completely different philosophy from classic packet: no connections, no handshakes, no guaranteed retransmissions. Packets are broadcast periodically-GPS position, short messages, weather data, telemetry-and anyone receiving them can use them, without the sender knowing whether they arrived. It is a system designed for wide-area information distribution, not for bidirectional communication. And precisely for this reason, it has endured: it is simple, robust, and works well even under difficult channel conditions.

There is a final chapter to this story worth mentioning, perhaps the most surprising: APRS has literally gone into space. The International Space Station has hosted amateur radio stations since 2000 as part of the ARISS program (Amateur Radio on the International Space Station), and among its operating modes is APRS packet. One of the onboard radios, in the Columbus Module, functions as a packet digipeater on 145.825 MHz. With just a simple VHF radio and a decent antenna, it is possible to send an AX.25 packet to the ISS as it passes overhead and see it reappear in the global APRS network, bounced from space.

Some amateur microsatellites also carry APRS transponders, further extending the network’s coverage.

A protocol born in the 1970s is still traveling through space today. It is hard to imagine a clearer demonstration of its longevity.

In summary

In amateur radio, “packet” is not just a block of data, but a concept that changed the way communication is understood: it introduces the idea of information split into autonomous units, enables reliable communication over noisy channels, and makes possible the creation of distributed networks without centralized infrastructure.

And above all, it represented-already in the late 1970s-the first real practical demonstration that a global wireless data network was not only possible, but could be built with relatively simple means. Today, classic packet radio is largely history, but it lives on in APRS and, even more importantly, in the technical legacy it left behind: a way of thinking about networks that has become the standard of the modern world.