Chapter 11. EDMAC CHANNEL
11.1 Theory Of Operation
As explained earlier, EDMAC is an acronym for
E
mbedded
D
istant-end
M
onitor
A
nd
C
ontrol.
This is a feature that permits the user to access the M&C features of modems that are at the
distant-end of a satellite link.
This is accomplished by adding extra information to the user’s data, but in a manner which is
completely transparent to the user.
On the transmit side:
The data is split into frames – each frame containing 1008 bits (except Rate 21/44 BPSK Turbo,
or when the data rates exceed 2048 kbps, where the frame length is 2928 bits, and Rate 5/16
BPSK Turbo where the frame length is 3072 bits). 48 bits in each frame are overhead, and the rest
of these bits are the user’s data. This increases the rate of transmission by 5% (approximately
1.5% for the Turbo BPSK cases, and for all data rates greater than 2.048 Mbps). For example, if
the user’s data rate is 64 kbps, the actual transmission rate will now be at 67.2 kbps. Note that the
user may also select EDMAC-2 framing, which uses a 2928 bit frame, and yields a 1.6%
overhead for all modulation types and data rates.
At the start of each frame, a 12-bit synchronization word is added. This allows the demodulator to find
and lock to the start of frame. At regular intervals throughout the frame, additional data bytes and flag
bits are added (a further 36 bits in total). It is these additional bytes that convey the M&C data.
When framing is used, the normal V.35 scrambler is no longer used. This V.35 approach is called
‘self synchronizing’, because no external information is required in the receiver in order for the de-
scrambling process to recover the original data.
The disadvantage of this method is that it multiplies errors. On average, if one bit error is present at
the input of the descrambler, three output errors are generated. However, there is an alternative
when the data is in a framed format: in this case, a different class of scrambler may be used – one
that uses the start of frame information to start the scrambling process at an exact known state.
In the receiver, having synchronized to the frame, the de-scrambler can begin its processing at
exactly the right time. This method does not multiply errors, and therefore has a clear advantage
over V.35 scrambling. This is fortunate, as there is a penalty to be paid for adding the framing: by
adding the extra 5% to the transmitted data rate, the effective Eb/No seen by the user will degrade
11–1