20.2 Signal Processing, Software & Codes
Key Takeaways
- SDR means Software Defined Radio—many transceiver functions (filtering, modulation, detection) implemented in software rather than fixed hardware
- DSP in a modern radio can control frequency, modulation, and detection, but does not control SWR (SWR is an antenna/feedline quantity)
- GMDSS DSC uses a 10-bit error-correcting code with data bits followed by a 3-bit error-correcting field
- SITOR-A/B characters are 7 bits with 4 zeros and 3 ones; Idle Signal (a) 0000111 is FEC Phasing Signal 1
- High-level data modes (Morse, RTTY/Baudot, ASCII, packet) and selective-calling codes sit on top of RF modulation—know which layer fails when symbols garble
20.2 Signal Processing, Software & Codes
Quick Answer: SDR = Software Defined Radio. DSP filters, demodulates, and can steer frequency / modulation / detection—but not SWR. DSC code = 10-bit error-correcting characters (data bits + 3-bit ECC). SITOR characters = 7 bits with 4 zeros + 3 ones. Idle Signal (a) = 0000111 = FEC Phasing Signal 1. Modes: Morse, RTTY, ASCII, packet ride on RF; selective calling addresses stations digitally.
Key topic 084 is the software and coding layer of Topic 3-M. If §20.1 explained how channels are shared, this section explains how bits and algorithms process the signal once it is digitized.
Software Defined Radio (SDR)
Definition
SDR means Software Defined Radio—not “software deviation ratio,” “SWR meter,” or “static dynamic RAM.” In an SDR, many functions that older designs built with crystals, discrete IF filters, and analog detectors are implemented as software (embedded firmware, FPGA fabric, or a PC host) operating on digitized samples.
| Traditional radio | Software-defined approach |
|---|---|
| Fixed crystal / mechanical filters | Configurable digital filters |
| Hardware modulator/demodulator per mode | Algorithms selected by software |
| Mode change often means hardware option boards | Mode change can be a firmware load |
| Alignment pots for every IF stage | Calibration constants and DSP parameters |
Why GROL techs still matter
SDR does not eliminate RF craftsmanship:
- The antenna, feedline, and front-end still determine noise figure, IMD, and SWR.
- Clock quality, ADC overload, and shielding still cause “DSP ghosts.”
- Illegal emissions still come from bad amplifiers and mixers—even if the waveform was “perfect” in software.
When a customer says “the software radio is broken,” separate RF path failures from firmware/configuration failures before swapping expensive boards.
Digital signal processing (DSP) — what it does and does not do
DSP samples an analog waveform (or processes already-digital baseband) and applies algorithms: filtering, noise reduction, demodulation, AGC, squelch logic, and modulation synthesis.
Pool fact: what DSP does not control
In a modern DSP radio, DSP can be described as controlling:
- Frequency (tuning / LO programming concepts),
- Modulation, and
- Detection (demodulation),
but DSP does not control SWR. Standing-wave ratio is a property of the antenna system and feedline match, outside the DSP core. No amount of clever filtering will turn a disconnected whip into a 1:1 load.
| Function | DSP involved? | Notes |
|---|---|---|
| IF/AF digital filtering | Yes | Notch, bandpass, noise reduction |
| Demodulation (AM/FM/SSB/digital) | Yes | Detection algorithms |
| Modulation synthesis | Yes | Waveform generation |
| Display of SWR from a sensor | Indirect readout only | Sensor is external to “DSP controls SWR” |
| Actual antenna match | No | Tuner, antenna, coax |
Filtering and demodulation literacy
Think of the classic receiver chain (Ch12) with a digital middle:
- RF front end still filters images and strong out-of-band signals.
- ADC digitizes IF or baseband.
- DSP implements sharp filters that would be bulky as analog crystal banks.
- Demodulator code recovers audio or bits.
- DAC (or digital audio link) drives the speaker or data port.
Advantage of DSP filters vs pure analog: flexible bandwidth, adaptive notches, and reproducible performance without aging inductors. Limitation: garbage-in from a saturated front end is still garbage-out.
High-level data modes: Morse, RTTY, ASCII, packet
Element 3 does not require you to send perfect Morse on exam day, but it expects you to place codes on the right “layer” of the system.
| Mode / code | What it is | Radio context |
|---|---|---|
| Morse (CW) | On/off keyed characters | A1A-class emissions; still used for identification and some backup links |
| RTTY | Radioteletype; historically Baudot 5-bit characters | FSK / AFSK over HF; ancestor of modern NBDP |
| ASCII | 7-bit (commonly stored as 8-bit) computer character set | Widest acceptance for computer-to-computer data exchange (Topic 083) |
| Packet | Framed packets with headers, payloads, error checks | OSI physical + link layers; AX.25-style radio packet thinking |
Troubleshooting map
| Symptom | Likely layer |
|---|---|
| No RF carrier / wrong frequency | Transmitter, LO, antenna |
| Strong RF but nonsense characters | Demod, shift, polarity, baud rate |
| Good characters but wrong meaning | Code set (Baudot vs ASCII), language tables |
| Intermittent frames, retries | Link-layer ARQ, multipath, collisions |
Selective calling codes — DSC and GMDSS digital calling
Digital Selective Calling (DSC) is the automated digital calling system used with GMDSS radios (VHF Ch-70, MF/HF DSC frequencies). Element 3 focuses on the character code, not the entire SOLAS carriage matrix.
DSC code structure (pool wording)
The code used for GMDSS-DSC transmissions is best described as a 10-bit error-correcting code starting with bits of data followed by a 3-bit error-correcting field.
Reject distractors:
- 10-bit code that starts with the 3-bit ECC then 7 data bits (order reversed),
- Simple 8-bit with one parity bit only,
- 7-bit code merely “transmitted twice.”
| Item | Pool fact |
|---|---|
| Total character length concept | 10-bit error-correcting code |
| Structure | Data bits then 3-bit error-correcting code |
| Purpose | Detect/correct errors so distress and routine digital calls remain reliable |
Selective calling in the larger picture
Selective calling lets a coast station or ship address a specific MMSI (or group/all-ships categories) without requiring continuous voice watch on every possible working channel. For the GROL maintainer:
- Verify MMSI programming and GNSS position input to the DSC controller.
- Confirm the Ch-70 or MF/HF DSC RF path (antenna, duplexers, dedicated DSC receivers).
- Remember: DSC is not the same as SITOR telex and not the same as NAVTEX—even though all are “digital maritime.”
SITOR character coding (bridge into marine NBDP)
SITOR (Simplex Teletype Over Radio) / NBDP (Narrow-Band Direct Printing) is the maritime telex family. Element 3 stores two hard coding facts under 084:
Character composition
The code used for SITOR-A and SITOR-B transmissions: each character consists of 7 bits with 4 zeros and 3 ones (constant-ratio code). Receivers use the 4/3 balance as a first error check.
| Mode family | Character rule to memorize |
|---|---|
| SITOR-A / SITOR-B | 7 bits: 4 zeros + 3 ones |
| Not the answer | 5-bit Baudot alone as the SITOR code description |
| Not the answer | 7 data bits + 3 separate ECC bits (that pattern is closer to other systems) |
Idle Signal (a) and FEC phasing
Which statement is true in the pool’s control-signal set? Idle Signal (a) (0000111) is used for FEC Phasing Signal 1.
Memorize the bit string 0000111 with the label Idle Signal (a) / FEC Phasing Signal 1. Do not invent that Signal Repetition or Control Signal 1 “is always” the FEC phasing answer when the stem asks for Idle Signal (a).
Codes and modes at a glance
| System | Element 3 coding takeaway |
|---|---|
| Computer data interchange | ASCII widest acceptance |
| GMDSS DSC | 10-bit ECC code; data then 3-bit ECC |
| SITOR-A/B characters | 7 bits, 4 zeros / 3 ones |
| SITOR FEC phasing cue | Idle Signal (a) 0000111 = Phasing Signal 1 |
| GSM voice compression | RPE CODEC (from §20.1) |
| P25 Phase 2 voice | AMBE CODEC |
| Packet structure | Physical TX; link framing |
Maintainer workflow: software radio + coded emissions
- Confirm RF health first — power, frequency, SWR, antenna. DSP cannot “software away” a bad match.
- Confirm mode and code — J3E voice vs F1B/J2B telex vs DSC; wrong mode yields perfect RF and unreadable traffic.
- Confirm identity programming — MMSI, answerbacks, EPIRB hex ID (survival gear in §20.4).
- Use built-in self-tests for DSC/controllers; avoid live distress alerts for curiosity.
- Document firmware versions on SDR platforms—field “regressions” are often config loads, not blown finals.
Exam-day checklist (3-M 084)
- SDR = Software Defined Radio.
- DSP: frequency / modulation / detection yes; SWR no.
- DSC: 10-bit error-correcting code; data bits followed by 3-bit ECC.
- SITOR: 7-bit characters with 4 zeros and 3 ones.
- Idle Signal (a) 0000111 → FEC Phasing Signal 1.
- Place Morse / RTTY / ASCII / packet on the correct functional layer when troubleshooting.
- Selective calling = addressed digital call (DSC), not voice shouting a ship’s name on Ch-16 (though voice remains the follow-on).
When software, DSP limits, and the DSC/SITOR bit rules are cold-memorized, Topic 3-M is complete. Topic 3-N applies the same thinking to marine VHF, MF/HF SSB, and SITOR operations.
What does SDR stand for, and which function does DSP not control in a modern DSP radio?
Which statement best describes the code used for GMDSS-DSC transmissions?
Which code is used for SITOR-A and SITOR-B transmissions, and which statement about SITOR control/phasing signals is true?
How should a GROL maintainer think about Morse, RTTY, ASCII, and packet when a “digital” maritime radio fails to deliver readable traffic?