Unformatted text preview: an be acquired using a sync marker defined in the information bit domain rather than the
encoded symbol domain, and detected after Viterbi decoding. This method relies on the fact
that frame sync is not required for successful operation of the Viterbi decoder but is
necessary for decoding the Reed-Solomon code. The Viterbi decoder is capable of finding its
own ‘node sync’ with or without the aid of known sync markers in the data stream. The
Reed-Solomon decoder has no effective method (other than trial and error) for determining
frame sync on its own, and so it must be presented with externally synchronized codeblocks.
It is irrelevant to the performance of the RS decoder whether this synchronization is
determined from the channel symbols or from Viterbi decoded bits.
In a similar way, the turbo decoder relies on being handed externally synchronized
codeblocks, but a bit-domain approach does not work effectively for turbo decoders, because
each constituent convolutional decoder is too weak by itself to detect a reasonable size
marker reliably, and because the powerful combined turbo decoding operation needs to know
the codeblock boundaries before it can iterate between permuted and unpermuted data
domains. Therefore, turbo code applications need to use channel-symbol-domain frame sync
methods as specified in the Recommended Standard (reference ).
Note that, for equivalent performance, channel-symbol-domain frame synchronization
requires longer sync markers and faster processing (at the channel symbol rate rather than the
Viterbi decoded bit rate). CCSDS 130.1-G-1 Page 8-6 June 2006 TM SYNCHRONIZATION AND CHANNEL CODING —SUMMARY OF CONCEPT AND RATIONALE 8.4 CERTIFICATION OF THE DECODED DATA (FRAME INTEGRITY
CHECKS) 8.4.1 GENERAL The CCSDS applications are packet-oriented, which means that data are collected and
transmitted in frames. With all coding options, and also for uncoded data, it is important to
have a reliable indication whether the decoded data is correct. A frame integrity check can be
used at the receiver side to validate the received frame or, when suitable, for requiring
retransmission in case of check failure.
As with the problem of randomizing the coded output, a universal solution to this data
validation problem exists in the form of a cyclic redundancy check (CRC) code, as specified
in the TM Space Data Link Protocol Blue Book (reference ).
8.4.2 DESCRIPTION OF THE RECOMMENDED CRC CODE Generally speaking, a binary CRC code is an (N,k) code obtained by shortening a cyclic
code, capable of detecting the following error patterns:
a) all error bursts of length N–k or less;
b) a fraction of error bursts of length equal to N–k+1; this fraction equals 1–2–(N–k–1);
c) a fraction of error bursts of length greater than N–k+1; this fraction equals 1–2–(N–k);
d) all error patterns containing dmin – 1 (or fewer) errors, dmin being the minimum
distance of the CRC code;
e) all error patterns with an odd number of errors if the generator polynomial G(D) for
the code has an even number of nonzero coefficients.
The circuits for coding and syndrome computation are simple feedback shift registers with
r = N–k cells.
In the CRC code used for the CCSDS TM Space Data Link Protocol Recommended
Standard, 16 parity check bits are added to every information frame consisting of (N–16)
information bits, according to the following generator polynomial:
G(D) = D16+D12+D5+1
Thus the rate of this CRC code is (N–16)/N.
The CRC circuit in the Recommended Standard (reference ) is preset to an all ‘1’ state
prior to encoding; this is a peculiarity of the CCSDS CRC, which implies that the 16 parity
check bits are inverted with respect to the usual CRC encoding (N–16) all ‘0’ information
bits generate 16 all ‘1’ parity check bits). CCSDS 130.1-G-1 Page 8-7 June 2006 TM SYNCHRONIZATION AND CHANNEL CODING —SUMMARY OF CONCEPT AND RATIONALE Serial concatenation of the CRC and a turbo code with nominal rate 1/3 is shown in figure 8-3. u cs
c2 C2 Figure 8-3: Turbo-CRC Encoder
The information frame is encoded by the CRC before entering the turbo encoder, and the
CRC syndrome is used to check the integrity of the decoded frame produced by the turbo
decoder at the receiver side.
8.4.3 USAGE CIRCUMSTANCES FOR THE RECOMMENDED CRC CODE The recommended CRC code is included in the Telemetry Frame and consists of 16 check
bits computed from the remainder of the frame contents. This code can reliably detect
incorrect frames with an undetected error rate of around 2–15≈10–5. This CRC code achieves
approximately the same undetected error rate for any of the recommended telemetry channel
A much lower undetected error rate is achieved when the RS code with E = 16 is used, either
by itself or concatenated with an inner convolutional code. In this case, the undetected error
rate of the RS decoder is on the order of 1/E!≈10–13, which is many orders of magnitude
better than the validation offered by the CRC code. Thus, the error de...
View Full Document
- Spring '14
- Information Theory, ........., Error detection and correction, CCSDS