Identification for Inverse Gaussian Channels

Abstract

We derive lower and upper bounds on the identification capacity of inverse Gaussian channels, a fundamental model for molecular communications in fluid environments. The analysis considers deterministic encoding schemes under a peak time constraint and characterizes the asymptotic growth of optimal codebook sizes. In the converse, by leveraging the Borel-Cantelli lemma and the super-exponential near-zero tail behavior of the inverse Gaussian distribution, we establish an almost-sure lower bound on the first-passage time of drifting diffusing molecules. This leads to the conclusion that the identification capacity exhibits super-exponential growth with the codeword length $n$, i.e., $\sim 2^{(n\log n)R_n}$, where $R_n$ is the coding rate.

18 pages, 2 figures; the converse proof has been extended to hold without restrictive noise assumptions