Exploring molecular distributed detection

Abstract

Chemical-exchange provides a fundamental mechanism by which microorganisms and biological cells communicate. Using this concept and the idea of biological sized nanomachines, this paper explores the detection of a undesired biological agent in a distributed setting. Distributed in the sense that the system information is dispersed across the nanomachines, each possessing limited communication capabilities. To study this problem, the molecular distributed detection system is divided into four major components. The first is the biological agent itself. The second is a collection of autonomous nano-sized sensors, capable of observing the environment, and releasing a certain type of communication molecule directly related to that observation. The third is an aqueous molecular transmission channel under drift, modeled using Brownian motion. The fourth is a fusion stage that collects the communication molecules and fuses this information to determine if a biological agent is present or not. Using this framework, we explore biological agent detection performance using optimal and suboptimal fusion rules in a parallel topology for diverse potential applications.