Fundamental to computer science is transmitting information using electromagnetic communication - the 0s and 1s of binary code. But nature's tiniest lifeforms have used a very different method for eons, and figuring out how they do it could revolutionize computers.
Bacteria make great use of what's known as molecular communication, in which they transmit and receive molecules to share information. One example of this is quorum sensing, where bacteria send around chemical signals to figure out their local population. This form of communication had long been ignored by information science, and it's only recently that the amount of information that can possibly be relayed using this method has been seriously considered.
Scientists at the University of Illinois have tackled this problem with a thought experiment. A transmitter sits in a fluid, replicating the conditions in which bacteria communicate. The transmitter emits a series of identical molecules, with the information encoded by the amount of time that elapses between each molecule released. In theory, that could be a fairly powerful form of communication, but the fact that the transmission is in a fluid introduces one hell of a randomizer: Brownian motion.
Brownian motion describes the apparently random movement of particles in a fluid. Indeed, this would work to completely scramble the molecules emitted; the molecules would arrive at the receiver at time intervals completely different from when they were first emitted - in fact, there's no guarantee they would even arrive in the right order - which would seemingly render the information unintelligible. At least, that was the prevailing wisdom.
Not so, argue the scientists. The uncertainty introduced by Brownian motion degrades the original information, to be sure, but not necessarily to a greater extent than the effect noise has on conventional communication. As long as the noise can be kept sufficiently small relative to the overall flow of information, the message will remain more or less intact. The same applies here, and the scientists point to factors like the fluid flow velocity and the molecular diffusion rate that can positively impact the flow of information and help reduce the relative "noise."
All of this is great news for computer scientists looking to add molecular communication as a complement to its electromagnetic counterpart. Certain computing problems are not well-suited to the highly ordered organization of the silicon chip, instead requiring something like the more disordered wetware that nature has developed. Indeed, there are certain types of computing problems that are perfectly suited to molecular communication - all we need to do is keep figuring out how to actually do it.