|
|
|
|
|
|
|
Radio Intercept R&S GA-082 Memory → ← Operation
Code programs
This page gives an overview of the various code programs provided by the
Rohde & Schwarz GA-082 FSK Analyzer.
It also lists the known USER program
packages and several optional classified B2 plug-in cartridges
for the B1 expansion unit. Whenever possible, the programs are
described in more detail further down this page.
Note that the contents of some cartridges are still secret.
|
|
The GA-082 can natively decode some well-known data formats,
that are handled by a set of built-in programs.
Confusingly, data formats are also known as modes (not to be
confused with operating modes).
The number of programs can be expanded by adding USER-packages in the
form of an EPROM-set (internal) or a plug-in cartridge (external).
There are four program types:
|
- Programs for code analysis (used in Mode 0 and Mode 2)
- Programs for decoding and displaying text (used in Mode 1)
- Programs for dealing with bitstreams
- Additional functions and special features
|
|
Each USER package may contain any combination of the above program
types. A USER package also determines the order in which the
internal and external programs are used to test an alien signal.
Only one USER package can be installed, and some packages
are restricted to a specific version of the system firmware.
From version 3.0 onwards, the number of additional programs was
greatly extended by adding the external B1 expansion unit.
When applicable, the program number is shown in Mode 2 as Nnn,
in which nn is the program number, e.g.
' N07 ' for
Baudot.
A program can be selected manually in Mode 2 after the device
has recognised an applied signal. It can also be selected
directly via the keypad by using the SELECT button,
e.g. SELECT-0-7 for BAUDOT.
Below is a list of all programs that are currently known.
The tick boxes at the right show what functionality is provided
by the program. The ticks have the following meaning:
|
| Analyser | This program does a measurement or checks for a particular condition. |
| Decoder | This program can decode the signal and produces the decoded text on the ANALYSIS display. |
| Function | This program provides a function, which does not necessarily produce any output on the display. It can be activated from the keypad by pressing SELECT followed by two digits. E.g. SELECT-88. Functions are executed immediately. |
| Bitstream | This program does not analyse the input signal, but allows decoding when selected manually. Research into this program type is still ongoing. |
| SECRET | These programs are classified codeword material (see below). ✱ |
| Collection | These items are in the Crypto Museum collection. ★ |
|
Some programs were developed by or for the German Armed Forces (Bundeswehr, BUND),
and are classified as SECRET codeword material (German: Schutzwort-VS
or SW-VS).
The cartridges containing these programs are marked GEHEIM-ANRECHT [1].
Knowledge about them was restricted (compartmented) to a particular user
group with the required security clearance and a need-to-know.
In the table below, these programs are marked with a ✱ symbol
in the column SECRET.
Please note that at present
we cannot (and will not) release copies of these programs in any way.
Programs N11-N19 are part of the USER10 package that was released
on system v2.x.
Programs N40-N46 are part of the USER20 package that was also
released under system v2.x. As the program names appear to be codewords, it might be
assumed that they are secret. For this reason they are marked with a
? in the table below.
Over the years, this package has been found by several people in legally
obtained surplus equipment. It is not marked 'secret' in any way.
All USER30 program cartridges are marked GEHEIM-ANRECHT.
It means that they contain compartmented material. Programs
N56 (ASCII), N68 (PACTOR) and N69 (AX25) however,
deal with commonly known communications formats of which detailed information
is available in the public domain.
Nevertheless we have chosen to respect the classification, and will not
disclose them.
|
|
The table below shows all programs that are known to date. The list is
assembled from the contents of the USER10,
USER20 and USER30 packages
that have been found, plus the contents of the
system firmware.
Items from USER13 are listed here under their
generic number, rather than the number they have in the
USER13 package.
It is believed that USER13 was an ad-hoc release.
|
|
|
|
| Analysis
| Decode
| Function
| Bitstream
|
| SECRET
| Collection
|
✔ = functionality, ✱ = secret, ? = unknown, ★ = in collection
|
|
|
N00
|
STOP-MOD
|
Permanent start or stop polarity (90% '1' or '0')
|
✔
|
|
|
|
|
|
★
|
|
N01
|
IDLE 1.1
|
Idle signal, 1 mark and 1 space sent alternately
|
✔
|
|
|
|
|
|
★
|
|
N02
|
IDLE 1.6
|
Idle signal, 1 mark and 6 spaces sent alternately
|
✔
|
|
|
|
|
|
★
|
| N03 | - | Unused | |
|
|
|
|
|
|
|
N04
|
IDLE 14
|
Idle signal with repeating arbitrary pattern of 14 bits
|
✔
|
|
|
|
|
|
★
|
|
N05
|
IDLE 28
|
Idle signal with repeating arbitrary pattern of 28 bits
|
✔
|
|
|
|
|
|
★
|
|
N06
|
IDLE 56
|
Idle signal with repeating arbitrary pattern of 56 bits
|
✔
|
|
|
|
|
|
★
|
|
N07
|
BAUDOT
|
ITA-2, async 7.5 bits (1 start, 5 data, 1.5 stop) 1
|
✔
|
✔
|
|
|
|
|
★
|
|
N08
|
ARQ-28
|
ITA-3, sync 7 bits (3 mark, 4 space, 4th char inv)
|
✔
|
✔
|
|
|
|
|
★
|
|
N09
|
ARQ-56
|
Same as N08, but every 8th character inverted
|
✔
|
✔
|
|
|
|
|
★
|
|
N10
|
ASY-ASCI
|
ITA-5, async 10-bit ASCII (7E1, 7E2)
|
✔
|
✔
|
|
|
|
|
★
|
|
N11
|
3 . 4
|
Check for 3:4 MARK/SPACE ratio
|
✔
|
|
|
|
|
|
★
|
|
N12
|
SITOR
|
Simplex ARQ Teletype Over Radio (single-channel)
|
✔
|
✔
|
|
|
|
|
★
|
|
N13
|
SIFEC
|
SITOR with Forward Error Correction
|
✔
|
✔
|
|
|
|
|
★
|
|
N14
|
PARITY14
|
Analysis of 12-data and 2-parity bits
|
✔
|
|
|
|
|
|
★
|
|
N15
|
ARQ-1A
|
ARQ-28/ARQ-56 with different alphabet
|
✔
|
✔
|
|
|
|
|
★
|
|
N16
|
SIEM-FEC
|
Similar to ARQ-1A but last bit inverted
|
✔
|
✔
|
|
|
|
|
★
|
|
N17
|
BAUER 10
|
AUTOSPEC, 10-bit data (5-bit error correction)
|
✔
|
✔
|
|
|
|
|
★
|
|
N18
|
SPREAD21
|
BAUER10 interleaved with 20-bits FEC data
|
✔
|
✔
|
|
|
|
|
★
|
|
N19
|
SPREAD51
|
BAUER10 interleaved with 50-bits FEC data
|
✔
|
✔
|
|
|
|
|
★
|
|
|
|
N21
|
MORSE
| Decoder for morse signals (disabled)
|
|
✔
|
|
✔
| |
|
|
|
|
|
N40
|
IKONE
|
? Possibly for analysis of Russian traffic
|
✔
|
✔
|
|
|
|
?
|
★
|
|
N41
|
UNSITTE
|
? Probably CIS-14
|
✔
|
✔
|
|
|
|
?
|
★
|
|
N42
|
WIMPER
|
?
|
|
✔
|
|
✔
|
|
?
|
★
|
|
N43
|
ANREDE
|
?
|
✔
|
✔
|
|
|
|
?
|
★
|
|
N44
|
BAUCHLAD
|
? Bauchladen (Tray)
|
✔
|
✔
|
|
|
|
?
|
★
|
|
N45
|
NACHTHUB
|
? Nachthaube
|
|
✔
|
|
✔
|
|
?
|
★
|
|
N46
|
GELAECHT
|
? Gelächt (Laughed)
|
✔
|
|
|
|
|
?
|
★
|
|
|
|
N54
|
STB 75
|
Philips SITOR-A (ARQ) and SITOR-B (FEC) modem
|
✔
|
✔
|
|
|
|
✱
| |
|
|
|
N56
|
ASCII
|
8, 10 and 11-bit ASCII (7G, A7G1, A7G2, A8G1)
|
✔
|
✔
|
|
|
|
✱
|
★
|
|
|
|
N61
|
MOERTEL
|
? Mörtel
|
|
|
|
|
|
✱
| |
|
N62
|
PARITAET
|
? PARITÄTEN
|
|
|
|
|
|
✱
| |
|
N66
|
WIESEL
|
VISEL / YUG-MIL / FEC-12
|
✔
|
✔
|
|
|
|
✱
|
★
|
|
N67
|
BIERABND
|
? BIERABEND
|
|
|
|
|
|
✱
|
★
|
|
N68
|
PACTOR
|
PACTOR-I
|
|
✔
|
|
|
|
✱
|
★
|
|
N69
|
AX25
|
(Amateur) Packet Radio
|
|
✔
|
|
|
|
✱
|
★
|
|
|
|
N78
|
PERIOD
|
Check for repeating patterns of 2-64 bits
|
✔
|
|
|
|
|
|
★
|
|
N79
|
M/S L...
|
Code statistics MARK/SPACE ratio etc.
|
✔
|
|
|
|
|
|
★
|
|
N88
|
|
Set Epson FX/MX-80 printer to Cyrillic (USER10)
|
|
|
✔
| |
|
|
|
|
N89
|
|
Set Epson FX/MX-80 printer to Latin (USER10)
|
|
|
✔
| |
|
|
|
|
 |
-
Improved BAUDOT program with measures the stop-bit length
and adjusts itself accordingly. The BAUDOT program can also
handle Cyrillic (Russian) characters.
|
|
This program is a data analyser. It does not produce any output
in Mode 1.
Between transmissions of asynchronous data, the signal is mostly
in one state; either 1 or 0. This program compares
the number of 1s against the number of 0s. When more than 90%
of the bits are 1s, the display shows N00 .
If more than 90% are 0s, the display shows -N00 .
STOP-MOD N00 More than 90% of the bits are 1s
STOP-MOD -N00 More than 90% of the bits are 0s
|
Between transmissions, the signal consists of alternating MARK and SPACE
bits (i.e. 1s and 0s). If this is the case for more than 90% of the signal,
the display shows:
IDLE 1:1 N01
|
Between transmissions, the signal has a 7-bit period of 1 MARK followed
by 6 SPACEs or vice versa. This is a typical idle state for BAUDOT signals.
A maximum error of 13% is allowed for each bit position. A minus sign is
shown before the program number if the signal consists of 1 SPACE followed
by 6 MARKs:
IDLE 1:6 N01
IDLE 1:6 -N01
|
|
The signal has a period of 14 arbitrary bits. The state of the individual
bits is irrelevant, as long as they are the same for each period. A maximum
error of 16% is allowed for each bit position.
|
|
The signal has a period of 28 arbitrary bits. The state of the individual
bits is irrelevant, as long as they are the same for each period. A maximum
error of 16% is allowed for each bit position.
Idling signals with a 28-bit period are typical for synchronous
error-corrected ARQ stations.
Inspection of the idling signal is possible by selecting Mode 6, 7 or 8
with a code length of 28 bits.
|
|
The signal has a period of 56 arbitrary bits. The state of the individual
bits is irrelevant, as long as they are the same for each period. A maximum
error of 16% is allowed for each bit position.
Idling signals with a 56-bit period are typical for synchronous
error-corrected ARQ stations.
Inspection of the idling signal is possible by selecting Mode 6, 7 or 8
with a code length of 56 bits.
|
Baudot is an asynchronous data format with a character length of 7 to 8 bits.
It is one of the most popular programs, as it is the original format in
which most FSK teletype traffic was sent. It is part of the native set of
programs that is provided by the system firmware of the GA-082.
It consists of an analyser and a decoder. The latter shows the decoded text
directly on the ANALYSIS display.
Each character begins with a start bit, which is a logical '0' (SPACE),
followed by 5 data bits, starting with the least significant bit (b0).
It ends with stop signal ('1' or MARK) that can be
1, 1.5 or 2 bits long. As the stop state ('1') is the same as the 'idle'
state, the stop bit can be stretched until the start of the next character,
as long as the minimum stop bit length is respected.
By default, characters are sent in the ITA-2 alphabet,
but other alphabets – such as Cyrillic –
are possible.
The initial version of this program (part of system firmware 2.0) has the
disadvantage that it only recognises the BAUDOT format with 1.5 stop bit.
For this reason, an updated and improved version was included in the
USER10 program package.
The updated version analyses the signal in more detail and can discriminate
between 1, 1.5 and 2 stop bits. The updated version was also included in
version 3.0 of the system firmware.
➤ More about ITA-2 and the baudot character format
|
|
Automatic Repeat Request (ARQ) is a format that is used for automatic
correction of low error rate signals. Synchronous transmission between
two stations requires two channels for each direction. Each character
is coded with 7 bits of the CCITT Alphabet No. 3 (ITA-3)
with a MARK/SPACE ratio of 3:4. A single bit error can be recognised
by a an incorrect MARK/SPACE ratio. In that case, a repeat request
is triggered through the return channel, resulting in a repetition
of the last three characters, preceeded by a RQ character.
The standard teletype characters are mapped as follows:
|
| Hex | 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | A | B | C | D | E | F |
| 0x | | | | | | | | I | | | | SP | | LF | E | |
| 1x | | | | R | | N | RQ | | | C | β | | D | | | |
| 2x | | | | L | | H | U | | | P | S | | A | | | |
| 3x | | O | FIG | | X | | | | LTR | | | | | | | |
| 4x | | | | G | | M | Z | | | V | α | | B | | | |
| 5x | | T | W | | Y | | | | Q | | | | | | | |
| 6x | | CR | J | | F | | | | K | | | | | | | |
| 7x | NUL | | | | | | | | | | | | | | | |
|
|
➤ ITA-3 alphabet
Control character
Printable character
ARQ character
A total of 35 characters can be created with a 7-bit word with a
3:4 MARK/SPACE ratio. Apart from the regular 32 teletype characters
(known from the ITA-2 alphabet), there are three
characters for continuous start polarity (α), continuous stop
polarity (β) and the repeat prefix (RQ).
The structure of single and multplex ARQ is described in
CCIR Recommendation 342-2,
which allows a wide variety of formats.
In its simplest implementation, a single channel is encoded
with every 4th character inverted (starting with the 1st character).
This results in a 28-bit period, as follows:
1234567123456712345671234567
Transmission of this sequence starts as soon as a station is powered on.
If no message is sent, it delivers a constant stream of idle characters,
which can easily be recognised in Modes 6-9. In Mode 1, the device
synchronises to the 7-bit character length and to the inverted 4th
character, and shows the decoded text on the ANALYSIS display
(without the (α), (β) and (RQ) characters).
|
This code structure is nearly identical to ARQ-28, except that
a period consists of 8 characters and that each 8th character
is inverted, starting with the first character. As follows:
12345671234567123456712345671234567123456712345671234567
|
This program provides basic support for asynchronous
ASCII encoded signals.
It is similar to BAUDOT, but is based on the
ITA-5 alphabet with
7 data bits for the character, plus one start bit, a parity bit
and one or two stop bits. The start bit is always a logical '0'
(SPACE), whilst the stop bit is '1 (MARK). The parity bit
is chosen such that the total number of 1s is always even.
The program, which is part of the native program set of the system firmware,
consists of an analyser that automatically recognises the format, and a
decoder that directly shows the readable text in Mode 1. The analyser also
recognises the inverted version of the format (-N10). A later version of the
ASCII analyser/decoder was provided by the N56 program,
which recognised a number of format variants, including synchronous ASCII.
➤ More about ITA-5 and the ASCII character format
|
This program is just an analyser.
It looks for 7-bit characters that have a 3:4 ratio between
MARK and SPACE bits. Such structures are used in ARQ, but also in many
other formats. The 3:4 ratio is used in systems that require a minimum of two
different bits between two adjacent characters. The program yields a
positive result when 80% of the characters in a 1024-bit block bit
have the expected 3:4 ratio.
It is likely that it also recognises a 4:3 ratio (shown as -N11).
An improved version of this program was provided as part of the
USER20 program pack.
|
|
SITOR, also known as SITOR A is a popular simplex ARQ system in which
two simplex stations alternatively use the same channel. It was
developed in the 1960s by the Dutch Post Office (PTT)
for use over HF radio. For maritime use
(ship-to-shore and vv) the operation is described in
CCIR Recommendation 476-3.
SITOR transmissions can be recognised
by a pulse signal with a period of 450 ms.
Below it is assumed that only one of the stations is analysed
with the GA-082.
|
| Hex | 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | A | B | C | D | E | F |
| 0x | | | | | | | | | | | | | | | | α |
| 1x | | | | | | | | J | | | | F | | C | K | |
| 2x | | | | | | | | W | | | | Y | | P | Q | |
| 3x | | | | β | | G | FIG | | | M | X | | V | | | |
| 4x | | | | | | | | A | | | | S | | I | U | |
| 5x | | | | D | | R | E | | | N | LTR | | SP | | | |
| 6x | | | | Z | | L | RQ | | | H | NUL | | LF | | | |
| 7x | | O | B | | T | | | | CR | | | | | | | |
|
|
➤ SITOR alphabet
Control character
Printable character
ARQ/SITOR character
SITOR uses a special 7-bit alphabet with 4:3 MARK/SPACE ratio 1,2
that is different from the ITA-3 alphabet
used with ARQ. The baudrate is always 100 baud, which means that
each bit has a duration of 10 ms. The ISS sends pulses of 210 ms,
which is equal to 21 bits or 3 characters.
ISS TX Information Sending Station
IRS RX Information Receiving Station
Master MSTR Station determining the time frame
Slave SLAV Station adapting itself to the Master time frame
The IRS sends 70 ms acknowledge pulses, by means of one of the
three special control characters (α), (β) or (RQ).
Alternating CS1 and CS2 characters notify the ISS of correct reception.
If the ISS receives the same acknowlege character on two successive
frames (i.e. 2 x CS1 or 2 x CS2), it repeats the last three characters.
Framing is different for Master and Slave, depending on which of the two
is the ISS. In the examples below, the upper line (red) is the Master,
whilst the lower line (blue) is the Slave. In example A the ISS is the Master.
In example B, the IRS is the Master.
Note the position of the
acknowledge character at the switch-over point of a station from ISS to
IRS. The Master places its acknowledge at the end of a time slot, whereas
the Slave puts the acknowledge at the beginning of the time slot.
In Mode 1, the decoded text is shown on the ANALYSIS display.
By default, control characters are not shown. To make them visible,
turn on the A-function on the keypad.
The control characters are the subsituted by the following
printable characters. The column 'Printed' shows the character that is
delivered to the V.24 port (printer or computer). The Column 'Display' shows
the character that is shown on the ANALYSIS display in Mode 1.
|
-
The USER10 manual [E] claims that SITOR has a 3:4 MARK/SPACE ratio.
This is incorrect. SITOR has a 4:3 MARK/SPACE ratio. In addition, the SITOR
character table presented in the USER10 manual (table 5.2) has the bits of all
characters inverted. In practice this should not be a problem, as the GA-082 can cope
with inverted formats. As a result, a minus sign (-) will be shown in front
of the program number in most cases: -N12.
-
In Table 5.2 of the USER10 manual [E], the codes for FIG- and LTR-shift
are erroneously swapped.
|
NOTE —
Make sure that the GA-082 stays at the selected polarity
when the station switches over from TX to RX. One way to ensure this
is to switch off the Fast- AGC of the intercept receiver. Set the IF bandwidth
to less than 300 Hz. In some cases it might be necessary to manually set the
shift of the GA-082 to 170 Hz and the baudrate to 100.
If synchronisation problems arise, turn on the CONST function
and set the time constant to 5.
SIFEC, the abbreviation of SITOR with Forward Error Correction, also known
as SITOR-FEC or SITOR-B, is the broadcast mode of SITOR.
It adds redundant data for error correction.
In the CCIR Recommendation 476-3 it is described as mode B.
Every second character is a repeat
of a previously sent character, with a distance of 35 bits (350 ms),
as illustrated in the diagram below.
SITOR B has two modes of operation, that are shown in the extra info field
in Modes 0 and 2:
COL collective mode
SEL selective mode
In selective mode, the data is inverted.
➤ More about SITOR
|
|
This program recognises groups of 14 bits that consist of 12 data bytes
followed by 2 parity bits. The parity bits are chosen such that a multiple
of 4 '1's (MARK) is counted in each group of 14 bits. This is achieved by
counting one of the parity bits twice. There are two variants:
|
- The last parity bit is counted twice
- The first parity bit is counted twice
|
|
This program was supplied as part of the USER10 program package.
It consists of an analyser only. Decoding was not possible at the time,
as the alphabet was unknown.
|
|
ARQ-1A is a 7-bit synchronous format that is very similar to ARQ-28
and ARQ-56, in that every 4th or 8th character is inverted. The only
difference is the alphabet, in which the characters do not have a 3:4 or
4:3 MARK/SPACE ratio. Instead it uses the 5-bit values of the ITA-2 alphabet,
shifted by one position. The least significant bit (lsb) is '0' for all regular
characters, and '1' for the three special ARQ-characters.
The most significant bit (msb) adds odd parity. As follows:
|
| | Character | ITA-2 | ARQ-1A | |
| # | Ltr | Fig | Hex | 543·21 | Hex | 7654·321 | Remark |
|
|
| 0 | NUL | 00 | 000·00 | 40 | 1000·000 | NULL, blank tape |
| 1 | E | 3 | 01 | 000·01 | 02 | 0000·010 | |
| 2 | LF | 02 | 000·10 | 04 | 0000·100 | Line Feed |
| 3 | A | - | 03 | 000·11 | 46 | 1000·110 | |
| 4 | SP | 04 | 001·00 | 08 | 0001·000 | Space |
| 5 | S | ' | 05 | 001·01 | 4A | 1001·010 | |
| 6 | I | 8 | 06 | 001·10 | 4C | 1001·100 | |
| 7 | U | 7 | 07 | 001·11 | 0E | 0001·110 | |
| 8 | CR | 08 | 010·00 | 10 | 0010·000 | Carriage Return |
| 9 | D | ENC | 09 | 010·01 | 52 | 1010·010 | Enquiry (Who?) |
| 10 | R | 4 | 0A | 010·10 | 54 | 1010·100 | |
| 11 | J | BEL | 0B | 010·11 | 16 | 0010·110 | BELL |
| 12 | N | , | 0C | 011·00 | 58 | 1011·000 | |
| 13 | F | ! | 0D | 011·01 | 1A | 0011·010 | Can also be % |
| 14 | C | : | 0E | 011·10 | 1C | 0011·100 | |
| 15 | K | ( | 0F | 011·11 | 5E | 1011·110 | |
| 16 | T | 5 | 10 | 100·00 | 20 | 0100·000 | |
| 17 | Z | + | 11 | 100·01 | 62 | 1100·010 | |
| 18 | L | ) | 12 | 100·10 | 64 | 1100·100 | |
| 19 | W | 2 | 13 | 100·11 | 26 | 0100·110 | |
| 20 | H | $ | 14 | 101·00 | 68 | 1101·000 | Currency symbol |
| 21 | Y | 6 | 15 | 101·01 | 2A | 0101·010 | |
| 22 | P | 0 | 16 | 101·10 | 2C | 0101·100 | |
| 23 | Q | 1 | 17 | 101·11 | 6E | 1101·110 | |
| 24 | O | 9 | 18 | 110·00 | 70 | 1110·000 | |
| 25 | B | ? | 19 | 110·01 | 32 | 0110·010 | |
| 26 | G | & | 1A | 110·10 | 34 | 0110·100 | Can also be @ |
| 27 | FIG | 1B | 110·11 | 76 | 1110·110 | Figures (Shift on) |
| 28 | M | . | 1C | 111·00 | 38 | 0111·000 | |
| 29 | X | / | 1D | 111·01 | 7A | 1111·010 | |
| 30 | V | ; | 1E | 111·10 | 7C | 1111·100 | |
| 31 | LTR | 1F | 111·11 | 3E | 0111·110 | Letters (Shift off) |
| |
|
|
ARQ characters
|
|
|
|
|
| 32 | RQ | | 07 | 0000·111 | Repetition |
| 33 | α | | 31 | 0110·001 | Idle signal a |
| 34 | β | | 49 | 1001·001 | Idle signal b |
|
|
Control character
In Mode 1, the printable characters are shown on the ANALYSIS display.
When printed via the V.24 port, the RQ character is shown as a 'r'.
The other ARQ characters (α, β) are not printed.
Characters with incorrect parity are reported as '*'.
These formats are recognised in Modes 0 and 2:
P28 Every 4th charcter is inverted
P56 Every 8th character is inverted
ID28 Signal is idling with 28-bit period
ID56 Signal is idling with 56-bit period
|
This is a 7-bit synchronous format with Forward Error Correction (FEC).
It uses the same alphabet as ARQ-1A, with the only difference that the most
significant bit (msb) is inverted for even parity. Furthermore, every
second bit of the signal is used for a convolutional error-correcting
algorithm, resulting in a total of 14 bits per character, as shown in the
diagram below:
The convolutional algorithm allows a blackout of several seconds to be restored,
as long as the structures before and after the blackout are undamaged.
The GA-082 does not take the convolutional algorithm into
account. Instead it only uses every other bit, starting with the first bit.
100000111111111000001111111110000011111111
The above sequence is the idle state. In this state, the GA-082 cannot determine
the baudrate. Text decoding in Mode 1 starts with synchronisation using the first
28 bits of the stream. After successful synchronisation, the decoded text is
shown on the ANALYSIS display. Erroneous characters are shown as spaces.
Too many errors will trigger a new synchronisation procedure.
|
This is a 10-bit synchronous format with Forward Error Correction (FEC).
The first five bits are identical to the 5 bits of the
ITA-2 alphabet.
The following five bits are a repeat of the first five bits if their parity is
odd, or an inverted repeat if their parity is even, like in these examples:
In Mode 1, single-bit errors are automatically corrected. Characters with two
or more bit errors are printed as space characters (SP).
The idle sequence is:
0000011011
|
SPREAD21 is a version of BAUER10 that is more resilient against burst
interferences. This is done by interleaving the bits of the 10-bit characters
with with 20 bits from other characters, with new characters starting every
10 bits. The bits are effectively arranged as a table with 10 rows and
21 columns as in the example below, with the bits of the first character
running from a1 to a0:
|----------- 10 bits ---------|
a1 y2 w3 u4 s5 q6 o7 m8 k9 i0 b1 z2 x3 v4 t5 r6 p7 n8 l9 j0 c1
a2 y3 w4 u5 s6 q7 o8 m9 k0 d1 b2 z3 x4 v5 t6 r7 p8 n9 l0 e1 c2
a3 y4 w5 u6 s7 q8 o9 m0 f1 d2 b3 z4 x5 v6 t7 r8 p9 n0 g1 e2 c3
a4 y5 w6 u7 s8 q9 o0 m1 f2 d3 b4 z5 x6 v7 t8 r9 p0 i1 g2 e3 c4
a5 y6 w7 u8 s9 q0 j1 m2 f3 d4 b5 z6 x7 v8 t9 r0 k1 i2 g3 e4 c5
a6 y7 w8 u9 s0 l1 j2 m3 f4 d5 b6 z7 x8 v9 t0 m1 k2 i3 g4 e5 c6
a7 y8 w9 u0 n1 l2 j3 m4 f5 d6 b7 z8 x9 v0 o1 m2 k3 i4 g5 e6 c7
a8 y9 w0 p1 n2 l3 i4 m5 f6 d7 b8 z9 x0 q1 o2 m3 k4 i5 g6 e7 c8
a9 y0 w1 p2 n3 l4 i5 m6 f7 d8 b9 z0 s1 q2 o3 m4 k5 i6 g7 e8 c9
a0 u1 w2 p3 n4 l5 i6 m7 f8 d9 b0 u1 s3 q3 o4 m5 k6 i7 g8 e9 c0
Character 'a' has a total distance of 9 × 21 = 189 bits.
Character 'b' starts at a 10-bit distance from character 'a'.
Character 'c' starts at a 10-bit distance from character 'b', etc.
As the data is spread over a long time span, there is a greater
chance of generating repairable 1-bit errors. In Modes 0 and 2, the GA-082
uses a 2048 bit buffer to process the signal. In Mode 1, the decoder uses
73 bytes of the signal past for synchronisation. Once synchronised,
it produces a character at 10-bit intervals. The reassembled characters
are further handled by the BAUER10 program.
|
|
SPREAD51 is similar to SPREAD21, except that the bits of each 10-bit
character are interleaved with 50 bits from other characters. This means that
the total distance of a single character is 9 × 51 = 459 bits.
As with SPREAD21, new characters start at 10-bit intervals.
In Mode 1, the decoder uses
133 bytes of the signal past for synchronisation. Once synchronised,
it produces a character at 10-bit intervals. The reassembled characters
are further handled by the BAUER10 program.
|
|
This program is a decoder for morse signals. It was found when analysing
the code of the USER10 program package,
but appears to be disabled. It cannot be selected manually.
|
IKONE (codename) (English: icon) is part of the USER20 program package
and probably also of the USER30 program suite.
The use of a codename suggests that it was probably classified.
It consists of an analyser and a decoder.
The exact purpose is currently unknown, but from analysis
of the code it seems likely that it was used for recognising and/or decoding
Russian traffic.
Real life signals delivered the following
on the display:
IKONE P2
|
UNSITTE (codeword) (English: bad habit) is part of the
USER20 program package and
also of the USER30 suite. The use of a codeword suggests that
it was probably classified.
It consists of an analyser and a decoder.
Test have meanwhile shown that UNSITTE is probably the same as CIS-14.
Live tests on a presumed Russian idling signal, delivered the following
on the display:
UNSITTE A N41
When the signal degraded, mainly due to fading, the display occasionally
showed the following:
UNSITTE A -N41
UNSITTE B N41
UNSITTE B -N41
|
|
WIMPER (codename) (English: eyelash)
is part of the USER20 program package
and probably also of the USER30 suite.
The use of a codename suggests that it might be classified.
The function of this program is currently unclear, but analysis of the
firmware shows that it consists of a bitstream processor and a decoder.
|
ANREDE (codename) (English: salution, preamble)
is part of the USER20 program package
and probably also of the USER30 suite.
The use of a codename suggests that it might be classified.
The function of this program is currently unclear, but analysis of the
firmware shows that it consists of an analyser and a decoder. Further
analysis of the firmware shows that the following (German) words can be
displayed or printed:
SEC AKT BAT ZIEL
This suggests that it might have been used to decode the message preamble
of some type of artillery traffic.
|
The program 'Bauchladen' (codename) (English: Tray)
is part of the USER20 program package
and probably also of the USER30 suite.
The use of a codename suggests that it was probably classified.
The function of this program is currently unclear, but analysis of the
firmware shows that it consists of an analyser and a decoder.
Further analysis shows that it can produce the following texts on the display
and/or the printer:
BAUCHL
BAUCHLADEN Q
|
The program 'Nachthaube' (codename) (English: Nightcap)
is part of the USER20 program package
and probably also of the USER30 suite.
The use of a codename suggests that it was probably classified.
The function of this program is currently unclear, but analysis of the
firmware shows that it consists of a bitstream processor and a decoder.
Further analysis suggests that this program was used for reading a particular
type of artillery traffic, as it can produce the following texts on the
display and/or the printer:
NACHTHAUBE
SYNCZAHL
STARTGR
ADRESSE
BTTR
ZIELNR
ZIEL/ARTTYP
LADUNG
ZUENDER
HOEHE
VISIER
RICHTUNGSFLAECHE
DISTANZ
GRANATENZEIT
KOMMANDOENDE
|
|
The program 'Gelächter' (codename) (English: Laughter)
is part of the USER20 program package
and probably also of the USER30 suite.
The use of a codename suggests that it was probably classified.
The function of this program is currently unclear, but analysis of the
firmware shows that it consists of just an analyser.
|
The program is an improved version of the
ASY-ASCI program (N10) that is present in the system firmware.
It can handle asynchronous as well as synchronous data
formats with 7 or 8 data bits, parity and one or two stop bits.
The program analyses the input signal and shows the recognised
format (Mode 0 or 2). It can also decode the data stream and show
the results on the ANALYSIS display in Mode 1. In Mode 2 is can
produce the following messages:
ASCI 11=A8G1 N56
ASCI 11=A7G2 N56
ASCI 10=A7G1 N56
ASCI 8 = 7G N56
ASCII NO N56
The N56 ASCII program was provided as part of the USER30 program suite,
that runs on version 3.x of the system firmware. When it is installed,
it takes precedence over the existing N10 ASY-ASCI program,
whenever one of its formats is recognised.
|
|
The program 'Mörtel' (codename) (English: mortar)
is part of the USER30 program suite
that runs on system version 3.x. The use of a codename suggests that it
might have been classified. No further details are currently known.
|
|
The program 'Paritäten' (codename) is part of the USER30 program suite
that runs on system version 3.x. The use of a codename suggests that it
might have been classified. No further details are currently known.
|
WIESEL (codename)
is part of the USER30 program suite that runs on system
version 3.x. It is belived to be classified and no details are currently known.
It is likely though, that it is identical to VISEL, YUG-MIL and FEC-12.
From analysis of the firmware, we know that it can produce the following
texts on the display:
WIESEL
WIESEL SUCHEN
WIESEL SYNC
|
BIERABEND (codename) (English: beer evening)
is part of the USER30 program suite that runs on system
version 3.x. It is a classified format (codeword: GEHEIM-ANRECHT).
No details are currently known.
From analysis of the firmware, we know that it can produce the following
texts on the display:
BIERABND
BIERABND SUCHEN
BIERABND SYNC
BIERABND SF=***
|
PACTOR was initially developed for Amateur Radio purposes, but is also
used by commercial parties. It is a synchronous simplex system with a fixed
timing cycle, that uses a two-tone FSK signal. It selects adaptively
between 200 and 100 baud, depending on propagation distortion.
PACTOR, also known as PACTOR-I, uses HUFFMAN data compression with
a factor of approx. 1.7 compared
to uncompressed ITA-5 (ASCII). Propagation disturbances may result
in loss of compression synchronisation and hence loss of data.
Detailed descriptions of PACTOR are available in amateur radio literature
and in the public domain. Note that the commercial implementation of PACTOR
is considerably different from the amateur radio version.
|
AX25 is the official name for Amateur Packet Radio. Although developed and
intended for use by Amateur Radio Operators, it is also used in commercial
applications. In addition, it was used by military groups on the Balkan
Peninsula during the conflicts in former Yugoslavia (1991-2001).
Detailed descriptions of AX25 (Packet Radio) are available in amateur radio
literature and in the public domain.
|
If none of the active programs tests positively with the input signal, the
GA-082 automatically invokes the PERIOD program (N78). It tests
the periodicy of the signal by searching for periodically occurring bits
with a period of 2 to 64 bits. The result is shown on the ANALYSIS display:
The first field shows the length of the period that has been discovered
in the range 2-64. The second field (i.e. the last 4 characters) shows the
signal Type. The following Types are known:
MARK periodic MARK bit
SPAC periodic SPACE bit
ASY periodic change MARK-SPACE
-ASY periodic change SPACE-MARK
IDLE periodic repetition of all bits
If no periodicy is found, the display shows just the current program
number in Mode 2: N78.
|
The STATISTICS program is the last one in the search sequence.
It can be called manually from the keypad (SELECT-7-9)
and always produces a result.
It analyses the last 1024 bits in the buffer and calculates the
MARK/SPACE ratio (M/S) and also the average number of adjacent bits
with the same polarity (L). The latter is obtained by dividing 1024
by the number of polarity changes in the 1024-bit buffer.
When the program is first activated the display shows:
As soon as 1024 bits of the input signal have been captured, a
calculation is made and the two values are shown as floating point
number at the position of the 
-characters.
With the built-in Baudot self-test (DEMOD-C-TEST) this program produces:
M/S=1.0 L=1.0BIT
As an example, an IDLE 1:6 signal produces these values:
M/S=.16 L=3.5BIT
|
|
This function does not produce any output on any of the displays,
but is executed immediately when the program number is entered.
It is part of the USER10 program package and
switches the connected EPSON printer (when present) to the Cyrillic
character set.
|
|
This function does not produce any output on any of the displays,
but is executed immediately when the program number is entered.
It is part of the USER10 program package and
switches the connected EPSON printer (when present) back to the
default Latin character set.
|
|
|
|
Any links shown in red are currently unavailable.
If you like the information on this website, why not make a donation?
© Crypto Museum. Created: Tuesday 28 January 2025. Last changed: Wednesday, 19 February 2025 - 15:18 CET.
|
|
|
|
|
|
|
|
| | |
