Converse Coding Theorems for Identification via Channels

In identification via channels, the sum of two types of error probabilities of identification goes to one as the block length of transmission tends to infinity at rates above capacity when channels satisfy some stochastic properties. This is well known as a strong converse theorem for the identification via channels. In this paper, we prove that the sum of two error probabilities tends to one exponentially and derive an explicit lower bound of this exponent function.