When instantiating a Viterbi decoding block, there are options to
specify
what the initial and final states should be or that they are unknown.
This
poses no problems if the FEC encoder is initialized similarly for each
block of data. However, when multiple blocks of data are encoded without
resetting the encoder to the same initial state each time, but instead
remains in the previous final state, there appears to be no option to
set
the Viterbi decoder initial state to its previous final state. As this
effectively increases the block size, and therefore improves
performance,
it seems that this would be a desirable feature. Are there any plans to
add
such a feature or are there other drawbacks that I may not be aware of?