On the error probability of a noisy channel network using network coding

We investigate the bit error probability (BEP) of a network using network coding (coded network). The network consists of independent binary noisy channels. An error marking algorithm is used to calculate the error weight (the number of erroneous bits) in the sinks. Hence, we can calculate the sink BEP of a coded network from the channel BEPs. Then we formulate the minimizing BEP problem. We show that coding schemes with the same flows may have different BEPs. Two approaches are given to reduce the complexity for the problem