Abstract
The memoryless additive inverse Gaussian noise channel model describing communication based on the exchange of chemical molecules in a drifting liquid medium is investigated for the situation of simultaneously an average-delay and a peak-delay constraint. Analytical upper and lower bounds on its capacity in bits per molecule use are presented. These bounds are shown to be asymptotically tight, i.e., for the delay constraints tending to infinity with their ratio held constant (or for the drift velocity of the fluid tending to infinity), the asymptotic capacity is derived precisely. Moreover, characteristics of the capacity-achieving input distribution are derived that allow accurate numerical computation of capacity. The optimal input appears to be a mixed continuous and discrete distribution.
Original language | English (US) |
---|---|
Article number | 6949028 |
Pages (from-to) | 2315-2329 |
Number of pages | 15 |
Journal | IEEE Journal on Selected Areas in Communications |
Volume | 32 |
Issue number | 12 |
DOIs | |
State | Published - Dec 1 2014 |
Keywords
- Additive inverse Gaussian noise
- Brownian motion
- average- and peak-delay constraints
- channel capacity
- molecular communication
ASJC Scopus subject areas
- Computer Networks and Communications
- Electrical and Electronic Engineering