Self-Organizing Sensory Systems

We consider a class of problems in the theory of robotic sensing systems. These particular sensing systems have many important applications in safety, surveillance, law enforcement, and defense operations, and have relevance even to biological life. The sensory systems we consider here are self-organizing, and therefore highly reliable. We will use a model problem to describe the properties of these systems and to illustrate application of the working theory, especially in relation to cellular automata theory. The main characteristic of this model is the absence of any central controller or clock. The sensory units are totally autonomous and the algorithms are fully distributed. Unlike cellular automata, the system is heterogeneous, kinetic and asynchronous. We introduce the computational model and present two protocols of self-organization. The protocols model the reconfiguration of the sensory system into (1) a high-resolution unidirectional sensing configuration and (2) a low-resolution bidirectional sensing configuration.