Figure 2 presents a snapshot of the simulation after 500 iterations (around 8 min).
After a certain number of PF iterations (around k > 6), however, the MAP estimates match the true source position better than the MMSE estimates.
For a larger number of iterations (around 60), dB gives the highest MI at dB, and dB gives the highest MI at dB. Figure 2 shows the BER performance of the two protection classes containing information bits for dB and dB.The BER of the non-UEP code is shown for comparison.
In our experiments, the typical number of outer iterations is around 200.
The algorithm was implemented in the programming language C. A single run with 2,000 iterations takes around three hours on a standard PC.
A conventional CVA-based decoder performs Viterbi iterations around the tail-biting trellis to find the optimal tail-biting path.
Then, iterations around this first value have allowed determining the optimum amplitude multiplication factor to minimize the residual noise on the difference between both channels.
In contrast, although sEM spends ∼40 iterations around G = − 0.70, a small jump that decreases energy at iteration 53 is followed by several iterations, which dramatically increase the energy.
They cycle involved has three parts which loop back around, prompting new iterations.
Due to the hierarchical scheme, the iteration converged around 28 iterations and yielded 0.8% maximum error.
