A problem of hypothesis testing on the crossover probability of a BSC is considered. We observe only the channel output and our helper only observes the channel input and can send us some limited amount of information about the input block. What kind of that information allows us to make the best statistical inferences? In particular, what is the minimal information sufficient to get the same results as if we could observe directly all data? Some upper bounds for that minimal amount of information and some related results are obtained.
Error Probability Dual Problem Channel Output Reliability Function Unique Root
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.
This is a preview of subscription content, log in to check access.