I understand that the message passing algorithm (Belief Propagation algorithm), when applied to a factor graph consists in an exchange in messages between the factor nodes and the variable nodes, these messages keep being updated until a certain number of iterations where we reach convergence. And here is my question: What is the number of iterations needed for the message passing algorithm to converge when applied to a tanner graph (factor graph)? Is there some conditions for the convergence to happen? if yes, what are these conditions? I am not using a loopy graph, my graph is an acyclic factor graph. Thanks in advance!
$\begingroup$
$\endgroup$
Add a comment
|