Lattice Semiconductor Turbo Decoder User’s Guide
7
Note: Signals are transferred on the rising edge. N is the block size.
Figure 5 shows the timing for data output. Once the Turbo Decoder has completed processing the data for a given
number of iterations it is ready to output the hard decision data. The decoder indicates this by asserting signal
rfo
.
Once the user detects this signal,
rfno
can be asserted to indicate that the output data can be received on the
output port
dout
. Thus for every
rfno
asserted the decoder will output the hard decision on the output port
dout
.
Once the decoder has output all data it will lower the
rfo
signal.
Figure 5. Decoder Output When Processing is Done
Note: Signals are transferred on the rising edge. N is the block size
.
Figure 6 illustrates how
inpvalid
can be used to control the flow of data into the decoder. The input data is
received in conjunction with input signal
inpvalid
. When
inpvalid
is asserted the input data on port
din
is
taken as valid data and written into the input data memory buffer. When the input port
inpvalid
is de-asserted,
the Turbo Decoder core ignores the value on
din
.
Figure 6. Input Data Handshaking with inpvalid Going Low
Note: Signals are transferred on the rising edge. N is the block size.
Figure 7 shows how
rfno
can be used to control the flow of data from the Turbo Decoder core. The output hand-
shake signals are
rfo
and
rfno
. Once the decoder is ready to give out the hard decision data it asserts
rfo
. The
receiving system will then assert the
rfno
signal when it is ready to receive the data. The decoder now puts the
hard decision data value on the port
dout
. When the decoder has output the last data it lowers
rfo
to indicate that
no more hard decision data is available in the decoder.
Lattice Semiconductor Turbo Decoder User’s Guide
8
Figure 7. Output Handshake with rfno Signal Going Low
Note: Signals are transferred on the rising edge. N is the block size.
When the decoder is processing data it will lower
rfi
if it cannot receive another data block. This occurs in two dif-
ferent scenarios. If only a single buffer has been chosen then the decoder will de-assert the
rfi
signal as soon as
an input data block is read in completely. In case of double buffer memory the decoder will lower
rfi
when it has
two blocks in memory and one of them is still being processed. Once the decoder completes processing the block,
it will assert both
rfo
and
rfi
. As shown in Figure 8, the average processing delay for the Turbo Decoder in the
case of 3GPP is 2.13xBxI, where B is the block size and I is the number of iterations.
Figure 8. Decoder State During Data Processing
Note: Signals are transferred on the rising edge.
Lattice Semiconductor Turbo Decoder User’s Guide
9
Parameter Descriptions
User-configurable parameters for each standard are shown in Table 1.
Table 1. User-Configurable Parameters
Parameter Name
3GPP 3GPP2
1
CCSDS
Decoder type
3GPP 3GPP2 CCSDS
Number of states 8 8 16
Maximum block size Default: 5114 Default: 20730 Default: 8920
Input data width
Range: 3-6
Default: 6
Range: 3-6
Default: 6
Range: 3-6
Default: 6
Window size
Range: 16 or 32
Default: 16
Range: 16 or 32
Default: 16
Range: 16 or 32
Default: 16
External memory
Range: Yes or No
Default: Yes
Range: Yes or No
Default: Yes
Range: Yes or No
Default: Yes
Double buffer
Range: Yes or No
Default: Yes
Range: Yes or No
Default: Yes
Range: Yes or No
Default: Yes
Decoder algorithm
Max Log Map (Default)
Log Map
Max Log Map (Default)
Log Map
Max Log Map (Default)
Log Map
External memory pipeline
Range: 1-6
Default: 1
Range: 1-6
Default: 1
Range: 1-6
Default: 1
Hard decision storage
Range: Yes or No
Default: Yes
Range: Yes or No
Default: Yes
Range: Yes or No
Default: Yes
1. 3GPP2 type is currently available only for LatticeEC™ and LatticeECP™ device families.

TURBO-DECO-O4-N1

Mfr. #:
Manufacturer:
Lattice
Description:
Development Software Turbo Decoder
Lifecycle:
New from this manufacturer.
Delivery:
DHL FedEx Ups TNT EMS
Payment:
T/T Paypal Visa MoneyGram Western Union