Digital Signal Types Revised 3/2016 Introduction Behold the humble radio wave. In order to communicate, people have come up with four ways to modulate or encode information onto it: 1. Turn it on and off, 2. Change the amplitude, 3. Change the frequency, and 4. Change the phase. Actually, there's kind of a fifth way, which is: 5. Do various combinations of 2-4, often on more than one wave. As digital modes evolve, we've seen some very clever ways of doing numbers 1-4, and #5 is coming up fast. --------------------------------------------------------- On-off keyed fuzzy modes A digital mode is a mode that's digital. This would seem to mean that it communicates in binary ones and zeroes, but radio is never quite that simple. There's a fuzzy third state which people understand better than machines. Fuzzy modes, then, typically use human intervention for the best copy. Some, like Morse code, require special training. Others, like FAX and Hellschreiber, don't. Morse telegraphy (CW, WT) Don't bury Morse yet. As a fuzzy mode, it can go farther into the noise than any other, though some of the multitone things are coming up fast. The fuzziness is what keeps Morse around. Early systems used mechanical printers to decode the dits and dahs, but people rapidly caught on that they could do better in their heads. It’s still that way. Computers, for example, can produce great copy on strong signals with perfect timing and no fading, but how many of those are there? Most of the time, people still do better. Morse code got a bad name when it was required for a ham radio license. Most countries don't require it anymore, but you still hear plenty of it on the bands. For several reasons, it remains a damn good way to communicate. Some people also consider it fun. It's an auditory experience, like playing a musical instrument. The best way to learn Morse is completely by ear. Treat it like folk music. Get into the sound and the flow of the code, and learn to become the code. When you can nod off at the radio, wake up, and realize you haven't missed anything, you're fluent at Morse. Morse code is a bit of a misnomer, because the radio uses something more resembling "Continental Code," though it's informally known as the "International Morse Code." Old-style Morse ("American Morse Code") is clattery, and several characters contain spaces. International Morse has more of a flow to it. International Morse uses the Latin alphabet, all caps, plus a few punctuation marks. Other "Morse codes" exist, for languages with a different character set, such as Russian. Still others expand the simple code in various ways, to communicate in more complex languages such as Japanese. Everyone's seen a telegraph key. It's a switch. Morse is either all the way on or all the way off, hence the name "on-off keying" (OOK). If the transition between on and off state is too abrupt, severe off- frequency interference called "key clicks" is generated. One can also key an audio tone and feed that to a single-sideband radio, as is done in most of today's ham transceivers. Today, Morse is copied with the receiver BFO on, just like sideband, and most people seem to like a tone around 700-800 Hz. Many radios use this offset for code, instead of the single-sideband 1.4 - 1.5 kHz offset. Some operators, myself included, have been known to lower the pitch for more punch when pulling a signal out of severe noise or interference. Of course, Morse can also be sent by FSK, which is heard sometimes, especially from some Russian military stations. One of the tones produces copyable code, while the other one produces an inverse that sounds like gibberish. The name "CW" actually means "Continuous Wave." It's an old name, from when various new devices with a steady oscillation were replacing spark gaps, which had a damped one. In the maritime service you'll also see it called Wireless Telegraphy (WT or W/T). Of course, there is very little Morse code left in this service, which has migrated to space-based and automated terrestrial systems. The Morse information rate is in words per minute (WPM), which is more relevant than simple baud rates, due to the varying element lengths and emphasis on timing. For counting purposes, Morse words have five letters. Text WPM is computed by sending the word PARIS. This works out to 2.4 times the rate in dots/second, as can easily be counted on automatic keys. At any normal rate and shaping, CW is extremely narrowband, making it possible to use sharp filters or DSP processors to pull signals out of sheer cacophony. This and the fuzziness of Morse make it still the all- time DX mode, working well below the noise floor. Hellschreiber People have a lot of fun with the name of this 70-year-old mode, which was invented in 1929 by a German named Dr. Hell. The name also means "Shiny Writing," from the silver drum once used. Originally, the transmitter resembled a crude sort of teletype unit, and the receiver was a primitive form of fax using paper tape. CW transmitters worked just fine, with the actual pixels of each character being sent using semi-synchronous, on-off keying. The receiver scanned each pixel onto a paper tape, printing each character twice for redundancy. The tape was then read by a human. Weak signals continue to print, but lighter. People do a pretty good job of interpreting the characters. That gave Hell a good reputation for doing better than RTTY on badly degraded HF circuits, and it was used quite a bit by the German military in WW II. Receivers were built by the Allies for intercept. Today, of course, it's emulated by computer. The most common mode is "Feld- Hell," "field Hell," which has a very distinctive, brrt-brrrt-brrrrrrrt sound. Today, this and some higher- tech variations are done with audio pulses fed to SSB transmitters, and received on a screen simulation of the moving tape. This is truly weird, but it's fun, and it's in most ham radio multimode programs. --------------------------------------------------------- Continuous-tone fuzzy modes FAX FAX stands for facsimile, in this case by radio. It's a very old mode, dating from wire telegraphy of the 19th century. FAX lends itself well to computer technology. While the ubiquitous office fax machine uses digital modulation, HF radiofax is still analog. Radiofax machines can be formidable affairs, with large drums, but computers work just as well in most applications. Being an analog mode, FAX will become fuzzy (noise) or smeary (multipath and delay), but still be recognizable to people. HF radiofax is actually frequency-modulation (FM), though tuned in standard upper-sideband (USB). Like audio FSK, fax can be sent by a USB transmitter merely by feeding in the correct tones. Deviation is 400 hertz up or down from 1900 hertz, low (1500) being full black, and high (2300) being full white. For this reason, FAX is tuned 1.9 kHz lower on USB radios. Computer sound cards handle the tones quite well, making FAX a popular way to send weather charts to laptop computers aboard small vessels at sea. Significant parameters are "drum speed," in lines per minute, and a rather complex thing called "Index of Cooperation," which is usually 576. Weather charts, being mostly white, are sent at 120 LPM, but the few remaining newspapers on HF (mostly from Japan) can go at 60 LPM. A radiofax usually has a band at one edge, causing a beep which gives away the speed instantly (and which, at 60 LPM, can sound like a time signal). The information rate of FAX is quite slow, with weather charts typically requiring 10-12 minutes for transmission. The Japanese newspapers, at their lower speed, take hours. Automated FAX uses dual-frequency autostart and end tones. As long as reception is good enough to receive these, a machine or computer can be set to grab hours of weather pictures while completely unattended. Analog Slow-scan TV SSTV is kind of souped-up FAX with color capability. It sends one frame at a time, and on HF it is best described as VERY slow-scan TV, really more of a photograph or computer picture exchange mode than any kind of television. Many of the same programs can decode FAX and SSTV, and also dump files off to either for transmission, using only the computer sound card. This makes SSTV a lot more fun than when people only held up photos to fuzzy cameras. SSTV started out like HF FAX, with frequency modulation and analog tones between 1500 and 2300 hertz. Monochrome cameras and converters like the Robot were used. Today, however, we see a truly bewildering number of color SSTV systems, with names like Martin, Scottie, or color Robot. All use a sync pulse at 1200 Hz, and send one line at a time with sequential information for the RGB colors. It's a fuzzy mode because, as long as it stays in sync, the picture just gets harder to see as the channel degrades. SSTV is also used on VHF and UHF, sometimes with repeaters, and even for automated pictures from spacecraft. All SSTV modes sound a lot beepier than FAX. Nowadays, one hears a lot about "digital SSTV." This is not really television, since the protocol used passes files as binary data, and these files could actually be anything. The most popular software for this mode is EasyPal, which error-checks data blocks and automatically requests repeats of missed ones. This works great for the receiving station, but not so good for people listening in, since you lose the whole picture after a certain number of errors. --------------------------------------------------------- Simple Teleprinting schemes These are all slightly different means of doing the same thing, namely sending a message as letters instead of words, for instant hard copy or storage at the other end, skipping the step of writing or typing it down. Teleprinters used to do just that - print. They would type everything out. Some pretty serious computing was done with automatic electric typewriters for output devices, though these were quickly replaced by line printers. Today, "printing" is usually loosely used to refer to any output of text, to a computer screen, teleprinter, or computer printer. HF teleprinting has a special set of parameters, in that bandwidth is precious, and radio paths are prone to noise, fading and phase distortion. All the different funny noises you hear are means of dealing with these limitations in different ways. As we've seen, machine CW and Hellschreiber were teleprinting schemes from the start, but most others were and are variations on radio teletype (RTTY). Coherent CW (CCW) This is an on-off keyed variation of Morse which is optimized for computer decoding. Sync is maintained by an idler sent in the absence of other characters. The original standard had a speed of 12 WPM. Since machines are doing all the work, it can be considered a very basic teleprinting scheme. QRSS Like coherent CW, QRSS is a variation of Morse code that is done by machines. In this case, though, the speed is very, very slow. It is sometimes measured in minutes per word rather than words per minute. The sound is that of a dead carrier with occasional brief interruptions. Reception is obviously by computer. QRSS is one of the weakest weak- signal modes going, capable of being copied from signals that are pretty much inaudible to humans listening on a speaker. The name comes from the international "Q" signal QRS, meaning "Reduce Speed." It's the ultimate QRS. --------------------------------------------------------- RTTY-like teleprinting systems AKA RATT (usually in the military) RTTY stands for Radio Teletype. "Teletype" was originally a proprietary tradename for mechanical teleprinters made by The Teletype Company. These were formidable machines, smelling of hot oil and requiring paper tape perforators for message storage. Most of the time, in fact, they were the same machines as used on wire lines. The military had thousands of them, feeding enormous transmitters or fed by diversity receiving systems that weren't much smaller. Whenever a new generation of Teletype machines was deployed, the old ones were given to amateurs, who'd sign pledges never to use the devices commercially. There weren't a lot of takers. RTTY was the lunatic fringe of ham radio, as these huge machines, and their associated tape devices, scopes, and general apparatus took over one's life, not to mention their living space, in a hurry. The revolution started with dedicated electronic devices, which tended to be expensive and hard to adapt to different modes. Otherwise, they worked very nicely indeed. Before long, though, computer software did the same thing, often for free, ultimately allowing any computer, tablet, or cell phone to receive RTTY nearly as well as thousand dollar dedicated boxes. Not that long ago, most of the odd noises on HF came from various teleprinting modes, all derived from the original RTTY concept. Now, not so much. Most of what is left is encrypted using any number of hardware-level schemes. However, a couple of stations in Europe still send RTTY weather bulletins. Plain old RTTY has two bit-states, called mark and space. Normal polarity signals have mark as the lower of the two, and reverse polarity signals have space lower. Some RTTY stations idle on mark, while others send pips at the character rate, which at least used to be called "diddle." Traditional RTTY uses frequency-shift keying (FSK), with two frequencies representing the two logic states. Most of the time, amateurs send tones made by computer sound cards to conventional single-sideband ham transceivers. While this is technically audio-frequency-shift keying (AFSK), the resulting on-air signal doesn't change in any audible manner. RTTY was traditionally tuned in lower sideband (LSB), though USB works just as well. In the computer era, with just about every other mode using USB most of the time, this is breaking down. Like FSK in general, both modes have a 100% duty cycle, meaning that the transmitter is at 100% output at all times. Amateur gear is rarely designed for this, and it usually needs to be run at half or sometimes even a quarter power. Here are some classic, RTTY-like systems: Baudot 5-bit asynchronous, aka ITA 2 (for International Telegraph Alphabet), oldest, still very widely used. A set of variations of ITA 2 exist for non-Latin characters, such as ATU-70 (Arabic). In common radio hobby usage, the terms "RTTY," "Baudot," and "ITA 2" are usually synonymous, with the other modes being described by name. This old digital code uses mark and space bits, with slightly longer pulses to frame characters. Originally, it was on-off keyed, but this was replaced by frequency-shift keying. Common shifts are 170 and 850 Hz, and common speeds are 45 and 75 baud. 50/250 is also heard. Note that the "baud" is an information unit equal to one digital event per second. While it's named for Baudot, the inventor of this code, most other digital modes also measure physical data pump speed in baud. As one finds out pretty fast with digital communications, a baud and a bit per second can be very different things. Beginners are advised to try amateur RTTY signals at 45 baud and 170 kHz shift, because they are generally, by law and custom, unencrypted. The ARRL headquarters station W1AW has regular schedules in RTTY and some other digital modes, with good, strong signals perfect for setting up software. Ham RTTY uses two sets of standardized tones. These are the "low tones" (1275 Hz mark, 1445 Hz space, 1360 Hz center), and the "high tones" (2125 Hz mark, 2295 Hz space, 2210 Hz center). This sounds a lot more complex than it is, because the common receiving technique is to simply set the computer to the audio center. 5 bits gives a restricted character set. Like Morse code, Baudot has no lower case. Unlike Morse code, it does have a FIGURES case for numbers and some other characters or signals, as selected with a SHIFT IN code, and deselected with a SHIFT OUT. Some languages such as Russian have a third, or even a fourth, shift for an expanded character set. If the SHIFT character is missed, the result looks rather strange indeed, and can result in loss of otherwise received data. Therefore, one can usually enable USOS (UnShift On Space) for tough situations. Baudot is asynchronous, with a distinctive chattering sound. Often, it idles on mark. The character pair "RY" contains all the possible states of the Baudot code, so one still occasionally sees RY strings sent for circuit testing or channel marking. These can be useful for checking your polarity, because if it's "wrong" you'll get RY as SG. Baudot's biggest problem is its lack of an error check, meaning that noise and fading cause missed or wrong characters. Some of the most amazing gibberish in all communication comes from degraded RTTY. Most later teleprinting modes come from different attempts to solve this problem. SITOR (Simplex Telex Over Radio) 4/7 bit error correcting synchronous Sitor was developed as an improved narrowband direct printing system for the maritime radio service. It's still used, though not as much. There are two variations. The first is Mode A, also called ARQ, Automatic Repeat reQuest, sent in 3-character blocks by rapidly alternating stations. The second is Mode B, Forward Error Correction (FEC), which runs continuously, and is used for broadcasts. ARQ sounds like nothing else on the bands. It gives a rapid chirp- chirp-chirp which is hard to miss. FEC sounds a little like speeded-up Baudot, only faster, higher pitched, and way less chattery. Both Sitor modes represent an early attempt at error-correcting RTTY. It also gives 100 baud, and more characters. In ARQ, stations take turns sending on a tight half-second cycle. The slightly longer burst is from the Information Sending Station, (ISS), and it contains characters with framing. The shorter burst is from the Information Receiving Station (IRS). The IRS error- checks and acknowledges received (ACK) or not received ("Negative acknowledgement," NAK). ACK/NAK signalling carried over into computer communication, and both signals are in the ASCII character set (See ASCII). The IRS resends NAK blocks, and if there are a lot of these, information throughput gets very slow indeed. Ultimately, it might just repeat the same block forever. Sync is everything in this mode. The timing is critical, and simplex demands transmit/receive switching that definitely strains the capability of amateur level gear. Long-path skip is impossible for the stations involved, and for everyone else it's just generally hard to monitor. However, the general concept of ARQ is fundamental to several newer systems. A handful of maritime coastal stations still take ARQ traffic. These send channel markers consisting of buzzes at the baud rate, followed by CW identifiers keying one of the tones. Only the Chinese ones ever seem to be especially active. A lot of the the remaining ARQ comes from the Egyptian Ministry of Foreign Affairs, for operator chatter before and after messages sent in a more modern mode. Most of this opchat is in an Arabic teleprinting alphabet called ATU-80, so software expecting the international standard prints gibberish. Also, even they seem to be phasing it out. FEC is simpler. It sends characters more than once, interleaved, and uses redundancy for the error check. Being a broadcast mode, monitoring it is a snap. NAVTEX Telex-formatted safety/weather in Sitor-B NAVTEX (Navigational Telex) is a special Sitor-B mode used to send Maritime Safety Information (MSI) bulletins on 490 and 518 kHz. The 518 ones are in English, while the 490 often use different languages. NAVTEX is an integral part of GMDSS, the Global Maritime Distress & Safety System, supplementing satellites in zones within range of coastal stations. Messages are given sequential headers, so that automated printers on board ship can skip ones previously copied. Stations repeat a 4-hour schedule, and vary for each of the world's navigation areas (NAVAREAs). The only ID they give is a sequential alphabet letter, but many lists of these exist, and one program in use here will even search these for you. AMTOR (Amateur Teleprinting Over Radio) Ham radio variation of SITOR AMTOR was once rather popular, but Pactor gives more bang for the buck, and it is a lot easier on equipment. Some ham radio programs copy AMTOR, meaning they'll get SITOR too. A variation of AMTOR uses an extended character set based largely on ASCII (see below). This makes AMTOR useful for bulletin-board systems and mailboxes, though these usually use packet radio or Pactor. VFT Voice Frequency Telegraphy Multiplex RTTY, often with orderwire Classically a multiplexed signal with 7 or more narrow FSK signals at once, using frequencies that fit into a voice channel. Sometimes the various subchannels send different messages, as in the old US military MULCAST, but the subchannels can also be used for time and frequency spreading, as in standard BR-6028. This gives robust error correcting on fading circuits. It sounds like a collection of intermittent, discordant beeps. Today's VFT usually has fewer channels. Versions with 2 and 3 have been logged coming from from at least the Swiss and Russian militaries. Other government/ commercial RTTY modes There are hundreds of these,in varying stages of obsolescence, but here are some: 36-50 (AKA CIS36-50, BEE, BEE-36) 50/200 FSK, used by Russian Navy 81-81 600-Hz shift FSK, used by Russian Navy ARQ-6/90 & 6/98 Single-channel 6-character block ARQ-E, ARQ-N Synchronous 5-bit full-duplex with framing and parity check ARQ-E3 Single-channel variant of ARQ-E with different character set, used mostly by French military ARQ-Mx (CCIR 242, CCIR342, etc) x-channel, time-division multiplex, 4/7 bit ARQ DPRK-ARQ (AKA 600/600) 600 baud, 600 Hz shift FSK mode used by the Democratic People's Republic of Korea ("North Korea") for diplomatic traffic. It also has a FEC mode. This is still heard quite frequently. DUP-ARQ Duplex ARQ used by Hungarian foreign service FEC-A 5-bit error correcting Baudot with parity FSK200 200 baud, 200/500/1000 shift FSK, used by Russian Intelligence HNG-FEC 15-bit scheme used by Hungary IRA-ARQ (BULG-ARQ) Fast, duplex ARQ with the IRA (ITA 5/ ASCII) character set, used by Bulgaria, Czech Republic, and Slovakia. POL-ARQ Special ARQ with 5-character blocks, used by Polish foreign service RUM-FEC 16-bit scheme used by Romanian foreign service SI-ARQx X=4 to 7 character blocks, different type of SITOR SWED-ARQ Variable block length ARQ scheme used by Swedish foreign service TWINPLEX 2-channel, 4-frequency-duplex, highly adaptive 100-300 baud mode used by UN, Interpol, and some foreign services. Seems to be giving way to Pactor in some of these places. --------------------------------------------------------- Computer direct printing modes Most new modes are generated directly by software running on computers. This can be done by sending audio to single-sideband transmitters, the same principle as AFSK, or by dedicated modems embedded in the radio itself. Modems ("modulator-demodulator" units) can use a single tone, which is then subjected to various complex modulation schemes, or multiple tones transmitted at once. The single tone can be modulated by FSK, phase-shift keying (PSK), or even amplitude. Many modems combine a number of these methods, and/ or use multiple phase states. The states are generally in multiples of 2, like 4, 6, 8, and so on. Multi-state signalling increases information density and/or redundancy, allowing better performance on badly degraded HF circuits. Multiple tones are frequently used, especially in PSK. One sees the number eight quite a bit, though it can run into hundreds at times. "Orthogonal" tones are frequently mentioned. This simply means that each tone decodes independently of the other ones. Automatic Link Establishment (ALE) does this. The current trend is toware ever larger numbers of tones, often in orthogonal frequency-division multiplex (OFDM), using extremely wide signals with sometimes hundreds of closely-spaced carriers. This stuff is cutting-edge, and very complicated. ASCII American Standard Code for Information Interchange Subset of ITA-5 A 7-bit, asynchronous, digital code, used mostly by certain types of computer networks. It offers a standard, 7-bit, set of 128-characters, derived roughly from ITA2, but with a lower case. There's a rather full set of control characters in the lower 32 values. Some of these come straight from old Teletypes, telling the terminal or computer to ring a bell (now more often a beep), or to feed a whole page. The most minimal ASCII terminal driver is still called a "TTY" by old timers. SHIFT IN and SHIFT OUT were carried over from RTTY, although not to implement a "numbers case." Most of the time, it was to provide some proprietary alternate character set which was not in the ASCII standard. In the dark ages of text-only computer terminals, these were often used for graphic elements, allowing the drawing of boxes, windows, and lines on the standard 80x24 text screen. ASCII can be sent via any kind of modulation allowing the two logical states of computer bits. The effective bit rate can be greatly multiplied by various esoteric modulation schemes, especially on telephone modems, while keeping a far more modest physical "data pump" baud rate. This is why the terms "baud" and "bits per second" are so universally confused. ASCII got kind of a bad reputation on HF. It started out as just FSK, at 110 or 300 baud, using telephone standards, but this did very poorly in bad band conditions. Today, it mostly refers to the standard character set itself, as used by many other modes. Various extensions to ASCII exist. Most common is 8-bit ASCII, which doubles the character set, allowing accented characters for such languages as Spanish and French, money symbols other than the dollar sign, and mathematical symbols. A newer standard, Unicode, builds on ASCII but adds many more characters. This standardizes all the "extra" character sets that existed, giving unique numbers to all the characters. Unicode is used quite a lot on the World Wide Web for exchange in different languages. It's truly amazing how many languages can work this way, especially if right-to-left typing can be enabled. ASCII characters are usually sent "framed," with extra bits to show start and stop, or for "parity," a crude error check. Common settings are 7E1 (7 bits, even parity, 1 stop bit) and 8N1 (8 bits, no parity, 1 stop bit). More complex framing schemes assemble characters into groups often called "datagrams" (see Packet Radio, below). MIL-STD-188-110A A standard for single-tone modems, often 8-state serial phase-shift keying, with a tone center of 1800 Hz, and adapted for use in many HF modem systems. A similar NATO standard exists, and the mode is fundamental to their common suite of HF waveforms. Other parts of the standard deal with parallel modems using 16 and 39 tones. Unlike the roughly similar sounding NATO STANAG 4285, 110A advises the receiving station of the correct circuit discipline. Both create a steady hiss, but 4285 runs a lot more continuously. Both offer long and short interleaves, and a variety of baud rates on a fixed symbol speed of 2400 baud. STANAG 4285 A single-tone, PSK, serial radio modem, again centered at 1800 Hz, with a bandwidth extending roughly from 300 to 3200 Hz. It's used mostly by NATO members, as a sucessor to RTTY. It has many throughput speeds, though the actual symbol speed remains fixed at 2400 baud. Most of the jet plane noises all over HF are made by this mode. The few unencrypted stations tend to use the same old ITA2 alphabet as RTTY, sent as 5N1, or even 5N2 to mimic RTTY's longer stop bit. They even retain the RYRYRYRY and QUICK BROWN FOX for testing, though these are no more relevant than any other character string. ASCII is also available, though I know of no stations audible here that use it. For listeners, this mode has the maddening aspect of revealing nothing about the proper circuit discipline from the sound or appearance of its waveform. It's all guess work. PSK31 This is 31-bps keyboard-to-keyboard teleprinting using binary phase- shift keying on a very narrow bandwidth, creating a warble centered near 1000 Hz. Clever design allows a 50 WPM throughput, and good DX. These parameters make the mode perfect for rag chewing between hams seated at computer keyboards. The full, 8-bit, ASCII code set is used, but the actual bit-level signalling is different. It is synchronous at the baud rate, and characters have variable lengths (Varicode). Thus can save a lot of channel time, not to mention creating an opportunity for more accurate decode. Since Morse code does this too, everything old is new again. The idle tone flips continuously at the baud rate, creating a piercing sound like an insect on the hottest night of August. It has a much lower duty cycle than all the FSK modes, but its efficiency means that lower power is usually used anyway. The mode really took off when computer sound programs allowed simultaneous decoding. Hams can tune to, say, 14070 kHz, set a wide receiver bandwidth, and print most of the stations inside. Clicking the mouse on the desired signal puts it front and center again. Nowadays, this mode also offers several higher baud rates, making for a broader signal, but higher speed. Some versions also use more phase states. MFSK8, 16, 32, 64 The name says it all: this is n-tone Multiple-Frequency-Shift Keying. It makes a really science-fiction multitone babbling noise, like running water or some kind of movie robot-speak. It's a good mode for HF, since it retains FSK's 100% duty cycle, eating power, but really punching through noise. MFSK has the interesting characteristic of decreasing its error rate as the number of states is increased, unlike phase-shift keying (PSK). Hams have really gone to town with MFSK. Variants have such names as OLIVIA, CONTESTIA, PAX, and others. It's even been adapted to send small images, like SSTV. The Voice of America's "VOA Radiogram" program does this every week, as a matter of course. PICCOLO One of several 6 and 12 tone sequential synchronous systems used by British and Australian government stations, alone or in VFT. Sounds musical, hence the name. 12-tone Piccolo allows the full ASCII character set to be transmitted. I think both modes have pretty much vanished from the air. COQ-8 Coquelet-8, a sequential 8-tone MFSK, mostly used at one time by Algerian embassies. Its name refers to the crowing of a rooster (le coq in French), giving an idea of its sound, which is similar to MFSK8. Other versions existed, with different numbers of tones, but today all these birds are approaching extinction. 36-50 (AKA BEE-36, T-600) FSK system used by Russian military. ALE Automatic Link Establishment (MIL-STD-188-141A, B (App. A), C (App. A)) AKA FED-STD 1045, STANAG 5066 An amazingly complex, 8-tone, orthogonal, PSK mode spelled out in several military standards, but also used by amateurs with free software. It's an 8 multi-tone PSK, instantly recognizable from its cyclic gobbling. The data is structured in "words" which also define its function. ALE benefits from its origins in military and government standards. These replaced a number of proprietary waveforms which only worked on one vendor's equipment, and the interoperability is what's made it perhaps the most used multitone PSK waveform in the world. ALE allows short text messages ("AMD") to be passed, but it is more of a networking protocol, enabling HF radios to choose the best frequencies and other parameters. Subsequent contact generally proceeds in another mode, such as voice or MIL-STD-188-110A. This is technically 2G ALE (second generation). A 3G ALE is specified in military standards (notably STANAG 4538), but its waveform is closer to 110A than what we're used to. It's heard on the bands, mostly from militaries. MT63 An amateur variation of the multitone PSK modes seen in military comm, using 63 data tones and one reference tone. In its most common configuration, it's a full kHz wide, and makes a buzzy roar often confused with jamming. Its wide bandwidth and multiple tones allow extreme spreading of both time and frequency, and several error checks, meaning excellent DX performance on badly degraded circuits. It's common to watch good copy keep coming from weak MT63 signals, or transmissions with other modes right on top of them. POCSAG I believe this stands for "Post Office Code Standardisation Advisory Group." POCSAG is far more common on frequencies above 150 MHz, though it's been heard in commercial use as low as 30 MHz. Its primary use was originally for radio paging, though it is currently being experimented with by amateurs. Believe it or not, there was a time when kids all walked around with beepers instead of cell phones. In its time, POCSAG probably initiated more drug deals than every other communication mode put together. It's basically framed, souped-up, RTTY, using direct FSK, as best taken directly from the FM discriminator on these higher frequencies. The mode became very sophisticated, with complex options for different levels of paging. FSK at this speed is extremely splattery, and super-powered paging transmitters used to be the worst interference sources on VHF/UHF. Happily, the coming of cell phones means that most of these have been taken out of service. FLEX Another paging-oriented transmission mode with more performance than POCSAG. Flex can handle a much greater volume, which was important back when everyone was carrying a beeper. --------------------------------------------------------- Computer data networking modes These use computer networking protocols similar to the ones seen in more familiar LAN and WAN technology. Most offer both automated and direct typing (keyboard-to-keyboard) modes, all using personal or laptop computers for terminals, most interfacing with radios through external controllers resembling more sophisticated modems. File transfers are often possible. Nearly all networking modes "packetize" data, which is usually 7-bit or 8-bit ASCII, but can also in some cases be raw binary - just ones and zeroes. This allows the network to handle different people's data at once by identifying, routing, and numbering packets, and it allows missed packets to be resent. The packets can be of varying lengths. They are exchanged, error checked, and processed by programs called "protocols," which provide standards for talking to other computers. The Internet is the most familiar packet-switched routing network, though it's hardly the only one. Internet needs an immense bandwidth, and it is usually implemented over telephone links and high-capacity fiber backbones (the "information superhighway"). However, there is no reason its underlying protocols can't be transmitted and received on the radio (see TCP/IP, below) Packet Packet-switched asynchronous data networking mode An ASCII protocol that sends data as little, framed "packets," one at a time, each a little message in itself. Data is broken down and assembled by a PAD (Packet Assembler/ Disassembler). The framing has several devices that make the receiving station process and print the data better. Among these is an error check of sorts. Some packets are for control, others are for information. Information packets are numbered, and can be received out of order. Long packets are often compressed, being sent and received as 8-bit binary data, in packets of varying length File compression is great for receiving stations, especially on slow HF circuits. It's not so great for listeners, who see the compressed data broken up into 8-bit characters as gibberish. Each packet is answered copy/no copy (ACK/NAK) by the receiver, and missed ones are resent. Packet thus slows down under difficult conditions, making file transfers a potentially excruciating proposition on HF. This has not prevented packet from doing some rather useful things, allowing e-mail, connection to remote "bulletin boards," and exchange of GPS positions over the air. Packet, like most networking systems, has a "link-layer protocol," also known as a "stack." This basically means that packet communication passes data from layers of software that interact with the user, down through layers that do various things, finally to the transport layer (the packaged data waveform), the physical layer (the transmitting/ receiving hardware), over the air, and then back up again. The highest layers, which would implement the kind of slick user interfaces that people are used to seeing on computers, are only now being worked out. A 2-way exchange of packet data begins with a connection. X.25 and AX.25 (amateur version of the standard), are the names of the most common system. AX.25 does badly on HF, with very slow throughput. It is heard daily on the amateur bands, however. Amateur packet network protocols are usually implemented in a terminal node controller (TNC). Until fairly recently, these were standalone devices that the computer talked to in text terminal mode. However, software TNCs are common today. Protocols have names like Aloha (developed at University of Hawaii, where else?), NET/ROM, TAPR (Tuscon Amateur Packet Radio) and KISS (Keep It Simple, Stupid). Where VHF packet started out with the 1200 and 2200 Hz "Bell Tones," HF uses a 200-Hz FSK shift. Straight packet is buzzy sounding, in short bursts,with a sync pulse at the beginning. Most AX.25 stations can be configured to "digipeat," relaying packets from other stations for sometimes thousands of miles. Another use of VHF packet is by hams to keep tabs on DX stations appearing on the HF bands, using a software system called "DX Cluster." Straight packet was originally seen as the salvation of ham radio, giving young computer geeks some incentive to get their licenses. There was quite a bit of interest at one time in creating a global network of digipeaters, but this lost out to the Internet, and packet radio became old news. It survives, however. One extremely interesting addition to packet radio is APRS, for Automatic Packet Reporting System. It does some rather neat things in the way of vehicle tracking, weather observations from the field, and mobile radio communication. In some ways, all this anticipated the smartphone apps that do similar things. Thankfully, though, its restricted bandwidth precludes selfies, memes, LOLcats, and farm games. Pactor, Pactor-I, Pactor-II, Pactor-III, Pactor-IV Packet Teleprinting Over Radio A composite mode; slowest one sounds a bit like Sitor Pactor is an evolving mode that was designed to combine the best features of packet radio (connections, file transfers, computer networking) with Sitor (robustness, simplicity). Hence the name Pactor, though the maker also says it's Latin for "the mediator." Whatever. Anyway, Pactor seems to add bells and whistles yearly, becoming ever faster and more complicated. Pactor's development has four levels, using Roman numerals. Pactor-I is FSK. Pactor-II is PSK or FSK, and Pactor-III switches between a dizzying number of modes on the fly. Pactor-IV is kind of new, and even more out there, producing some really strange sounding signals at times. Everything higher than Pactor-I is very proprietary to SCS, a European firm started by hams. They offer it in several very nice, and very expensive, external modems. They've licensed it out to a handful of other companies, usually for modems or decoder packages costing more money than the radios they run on. Pactor-I is mostly kept around for identification, and the initial callup and connection. I can snag an occasional call sign this way. All levels incorporate robust error-checking to overcome the problems of HF packet. Pactor is becoming the mode of choice for the many amateur and commercial HF e-mail systems, which are pretty much the main use of the maritime NBDP frequencies at this time. They provide an attractive alternative to satellite phones. Typical networks include SailMail, Bushmail, Winlink, Airmail, and a number of others. As Pactor speeds up, its bandwidth increases dramatically. It pushes the limits of what can fit in a voice channel, but then so do 110A and 4285. Despite its expense, Pactor is a well thought out system, and it seems to be winning in the marketplace. WINMOR WINMOR came out around 2008-2009, originally billed as a an affordable amateur alternative to Pactor for the Winlink wireless e-mail network. It seems to have two modes: narrow (500 Hz with two 46.875 baud 4FSK tones), or wide (1600 Hz, with eight 93.75 baud 4/8/16 state PSK tones). WINMOR is usually ARQ, but it also has an FEC broadcast mode. CLOVER Phase-shift keyed, ARQ, duplex, 4-tone, radio modem This mode was invented by hams, and remained proprietary to the HAL company for HF data transfer. Clover-I makes an intermittent bleeping sound. Clover, like Pactor, changes modes on the fly to pick the fastest one the circuit will support. Unlike Pactor, though, most of these modes remain in about the same bandwidth. The current version, Clover 2000, uses various combinations of PSK and QAM, typically with 8 tones. It's been around quite a while, but it's still in use by several agencies. G-TOR Golay Teleprinting Over Radio, named from a type of data coding used. It was developed by Kantronics, a maker of packet gear, and it remains proprietary to them. It always seemed liked a well thought out system, adding some of the bells and whistles in military standards to amateur- level equipment. By all descriptions, it was fast and robust. All the information available through Google searches seems dated, and the status of G-TOR cannot be determined at this time in 2016. TCP/IP Transfer Control Protocol/Internet Protocol Packet radio use of the same protocol suite used on the Internet, as opposed to X.25. Once everything's working, the user interface is quite a bit like the old dialup unix days, with use of such programs as telnet and ftp. It is even possible to use Windows. It is also possible to use TCP/IP via AX.25, using "datagram" mode. TCP/IP is useful for "gateway" stations, which allow HF radios to connect to the Internet, to military IP routed networks, or to e-mail services that seamlessly connect to the Internet and other packet protocols. HFDL High-Frequency Data Link HFDL is a very slick, global system deployed by ARINC (division of Rockwell Collins) worldwide for automated communication with aircraft. It works a bit like packet radio, and a bit like MIL-STD-188-110A, only using variably offset time slots that are synchronized throughout the entire system. The many available frequencies are kept in a "system table" which coordinates their use worldwide. The transmissions are carefully timed single tone PSK centered on 1440 Hz, with a fixed 1800 baud symbol speed and a sync beep at the start. Effective baud rates are changed in the usual manner by adding or subtracting the number of phase states. Every 32 seconds, the ground station identifies with a "squitter" containing its ID, the frequencies in use (as table numbers), and the allocation of time slots. The other slots pass very comprehensive data structures from aircraft (downlinks) or to aircraft (uplinks). These are parsed into various data units and decoded. The amound of information that can be passed over HF radio in a few seconds is downright astounding. Most aircraft report their position, allowing a number of amateur plotting programs to keep track on maps. HFDL can also pass tightly formatted text messages known as HF ACARS. This is a modified version of the Aircraft Communications, Addressing, and Reporting System deployed on VHF some years back by ARINC. Some of these are weather data or even company chatter, which is interesting. All this is completely seamless to the customer, who connects to a "GlobaLink" using HF, VHF, and satellites. Link-11 AKA TADIL-A, Alligator The famous US Navy/NATO "gator," used for many years to share radar tracking data between participating units and a Combat Information Center (CIC) aboard a ship. There are two waveforms. CLEW (Conventional Link Eleven Waveform) is an intermittent system with 16 tones. The lowest one is unmodulated, and used for Doppler lock. The others are for data, with the highest one also providing sync information. All but the Doppler tone are DQPSK (Differential Quarternary Phase- Shift Keying.) The Doppler tone can be heard as that distinctive bong just before the gator hiss. The time sync scheme in use makes a a rhythmic sort of sound mixing long and short bursts, sometimes in pleasing, percussive beats. The "alligator" name comes from the hissing sound of the actual data transmission, which greatly resembles a noise that gators emit when they are happily making little gators. The other waveform is SLEW, Single-Tone Link Eleven Waveform. As the name implies, it's one tone using 8PSK at 2400 baud, plus the Doppler tone. It sounds roughly similar. SLEW is said to work with skip propagation, while CLEW is strictly ground wave. As should be obvious for this application, the information payload is deeply encrypted. Link-11 has been around for quite a while now, but it's still all over the bands. Often it appears in double-sideband mode, usually with the same signal in both. It is one of a whole suite of TADILs (Tactical Digital Information Links). It is said that Link-11 is due for replacement by a newer standard called Link-22. DSC Digital Selective Calling DSC was implemented as part of the Global Maritime Distress and Safety System (GMDSS) which replaced the other things we used to hear in the maritime mobile service. Superficially, it sounds like Sitor-B, using a 170-Hz shift at 100 baud. However, it is basically a calling mode sent in short bursts which are evaluated as 10-bit characters where 3 are parity bits. Automated DSC calls come in several types. The most common ones seem to be position checks and safety tests. These are usually between a ship and a shore station, which is usually connected to a Rescue Coordinating Center (RCC). Often, though, a ship will actually call another one in this mode, specifying a frequency and mode for further communication. The call signs used are number sequences known as MMSI, for Maritime Mobile Service Identity. Lists of MMSI are available online, allowing the identification of vessels and coastal stations. Many computer programs do this lookup automatically. --------------------------------------------------------- Other Modes This will always be incomplete, as new ones come out yearly. 39-tone (AKA NATO, HARCO-39) Descriptively named military standard, now also available for civilian use, with 39 tones, allowing redundancy for error tolerance under fading conditions. It's seldom that fades hit all tones equally. The sound is very distinctive, with a sync tone followed by all 39 coming up, followed by what sounds like white noise. (It's an axiom that, as information density increases, any system will sound more like noise.) Variations include ANDVT, the military's Advanced Narrowband Digital Voice Terminal, which is not a scrambling scheme but can and usually does interface with secure voice modules. Other federal and military terminals can securely transfer files. The encrypted mode takes audibly longer to synchronize. DGPS Differential Global Positioning System DGPS is 100 or 200 baud minimum-shift keying (MSK). It's optimized for narrow bandwidth on the low (283.5 to 325 KHz) frequencies used. The 200 baud is twice as wide as the 100. The purpose of DGPS is to improve GPS accuracy. Its transmitters are at known and unchanging locations, giving a reference against the ever moving satellite constellation when used by receivers with this capability. Out in the flats, using the "hockey puck" receiver and DGPS, I can get fixes that tell what lane of traffic I'm in. There are something like 63 message types, but by far the most common one is corrections. These are sent as data and decoded at the receiver, but there is another message type that allows straight ASCII. At least here in California, the DGPS band is becoming packed. Just about every possible frequency has a signal on it. It's obviously a very well used mode. Digital Slow-Scan TV Digital slow-scan TV is kind of a misnomer, because "under the hood" it's basically a file transfer, with features relating to images layered on top of it. Also, like analog SSTV, it's really about still pictures more than television. Like analog SSTV, the goal is to transfer a single image at a time. The images tend to be about the same size as analog SSTV pictures. The quality, however, is greatly improved, sometimes dramatically so. The current standard computer program for doing these transfers is EasyPal, by VK4AES. It uses a waveform greatly resembling a much more narrow version of the Digital Radio Mondial (DRM) used in some HF broadcasting. The underlying modulation is QAM, Quadrature Amplitude Modulation, which achieves a higher information density by encoding data in both amplitude and phase. QAM 4 seems to work best on HF. DRM is 100% duty cycle, requiring power reduction, and then its average power is fairly low, reducing efficiency further. It is definitely a strong-signal mode. Worse, digital SSTV is far harder than analog for third parties to monitor, because they can't ask for repeats of bad blocks the way the receiving station does. Japanese Slot Machine This odd noise gets its name from its syncing idler, which repeats a jangly, semi-musical tune sounding exactly like a gambling machine that was common at the time. It was quickly DF'd to Japan, and the best guess anyone can make is that it's used by the Navy to send deeply encrypted messages in brief payloads that sound like a big hiss. Again, it's axiomatic that the denser a signal gets, the more it sounds like noise. The underlying modulation is QPSK, but I/Q displays show that one phase state vanishes at times, helping the idler get its distinct sound. The entire signal is chopped up with a clicky interruption about 11 times per second. This appears to delimit data frames. In general, this thing is completely unique. There's nothing else even close to it on HF. JT65 JT65 is one of a whole suite of FSK modes designed for various weak- signal propagation types. The first software that incorporated it seemed to be intended for Earth-Moon-Earth ("Moonbounce") on frequencies above HF. Now, however, it appears in a program designed for HF use, eponymously named JT65-HF. The modulation mode is slow MFSK with 65 tones. The "JT" stands for Joe Taylor, W1JT, who designed it. Messages are sent compressed and using FEC for an error check. They are by necessity short, usually a call sign and very short text message. These take about 50 seconds to send, and sound like a slow warble. As with many JT modes, time sync is critical. Messages begin shortly after the 0 second of every UTC minute, and the end of the minute is for decode. The decoder can handle more than one message at once, and it is very robust. The result is yet another mode that can work well into the noise floor, to the point of near inaudibility to the ear. Listening in on JT65, which is usually around 76 kHz above the lower ham band edge, is easy and quite productive due to the mode's extreme DX capabilities. My best is South Africa on 10 meters, from a station using very low power. That's about 11,000 miles! Mazielka AKA X06 (Enigma) 6-tone Russian selcall, where MFSK tones stand for figures 1-6. It repeats a single 5-figure group. It is used by the MFA. Usually this is a callup before messages in another mode, usually Serdolik, and usually on a different frequency. RDFT Redundant Digital File Transfer RDFT is an old mode, as these go, originally intended for digital SSTV over ham radio, but now used only by the Cuban intelligence service to broadcast encrypted messages in the form of binary files introduced by the traditional digitized voice. These are easy to hear in the continental US and Europe, and have been designated HM01 by the "numbers" sub-hobby. The waveform consists of 8 hissy PSK tones, framed by multiple pure audio tones of various frequencies. It's known from earlier Cuban trials that the program being used is DIGTRX, a sort of precursor to EasyPal. People who can find this program and get it to work can decode the Cuban files just fine, though all you get is a bunch of encrypted binary data. Been there, done that. Selcal Selective Calling The selcal scheme described here is used on HF for the activation of HF receivers in the cockpits of airliners on long oceanic flights. It is a means by which a flight crew can keep a channel open for calls without having to listen to HF noise. Selcal is a sort of slowed-down MFSK. It transmits two tones for one second, then a short pause and two tones for another second. There are 16 of these, corresponding to the letters A, B, C, D, E, F, G, H, J, K, L, M, P, Q, R, and S. The demodulated result is two pairs of letters such as AB-CD. These letters are assigned to airplanes near the beginning of their service life, and tend to remain static regardless of the plane's owner. There are more airplanes needing this service than combinations available. Most selcals go to more than one plane, though an attempt is made to assign them to aircraft operating in different parts of the world. Since big jets go everywhere, this doesn't always work. Selcal is tested when an airplane checks in with a ground station, to make sure it's right and that it works. It's most often heard when a ground station takes a request from a plane under oceanic air traffic control for a different altitude. These have to be sent to ATC by the operator, and there's always a delay. The selcal is sent when the reply is ready. There are other types of selcals out there on HF, but the airplane one is the most commonly heard. Serdolik AKA Crowd-36 This is 34-tone MFSK at 40 baud, used by the Russian MFA. The name comes from a type of Russian carnelian. It can be announced by a selcall in "Mazielka." WSPR Weak-Signal Propagation Reporter AKA "Whisper" This is another ultra-weak-signal mode from K1JT. It sends short messages very slowly, using what is essentially very narrow MFSK4. To the ear, it just sounds like a slightly wobbly carrier. Real communication can take place well below the noise floor, at the point where the signal is barely audible, if at all, by speaker. These messages contain 50 bits of information. It requires one minute and 50.6 seconds to send these. They are automatically generated, and usually contain the amateur call sign, Maidenhead grid locator, and a dBm figure relating to power. Timing is critical. Messages start one second into every UTC minute, and the more exact, the better. Frequency measurement is also critical, to whatever the limit of the transceiver allows. The user is instructed to compute the ppm error for each amateur band using WWV as a reference. The WSPR program can be configured to upload received messages, plus time and frequency data, to a central database of spots which is effective in building records of propagation over time. # # #