Cryptography Reference
In-Depth Information
Application
Turbo
code
Termination
Polynomials
Rates
CCSDS
(deep space)
binary,
16 states
tail bits
23, 33, 25, 37
1/6, 1/4, 1/3, 1/2
UMTS,
CDMA2000
(mobile 3G)
binary, 8
states
tail bits
13, 15, 17
1/4, 1/3, 1/2
DVB-RCS
(Return
Channel on
Satellite)
double-
binary, 8
states
circular
15, 13
from 1/3 to 6/7
DVB-RCT
(Return
Channel on
Terrestrial)
double-
binary, 8
states
circular
15, 13
1/2, 3/4
Inmarsat
(M4)
binary,
16 states
none
23, 35
1/2
Eutelsat
(skyplex)
double-
binary, 8
states
circular
15, 13
4/5, 6/7
IEEE 802.16
(WiMAX)
double-
binary, 8
states
circular
15, 13
from 1/2 to 7/8
Table 7.2 - Standardized applications of convolutional turbo codes.
formance is not very sensitive to permutation, on condition of course that
it respects at least the principle of dispersion (which, for example, can be a
regular permutation). For a low or very low targeted error rate, performance
is dictated by the minimum distance of the code and the latter is highly
dependent on the permutation Π .
d- The puncturing pattern must be the as regular as possible, in the same way
as for classical convolutional codes. However, it can be advantageous to
have a slightly irregular puncturing pattern when we are looking for very
low error rates and when the puncturing period is a divisor of the period of
the generator polynomial of recursivity or parity.
Puncturing is performed classically on the redundancy symbols. It can be
envisaged instead to puncture the information symbols, in order to increase the
minimum distance of the code. This is done to the detriment of the convergence
threshold of the turbo decoder. From this point of view, in fact, puncturing
Search WWH ::




Custom Search