Trail Systems as fault tolerant wires and their use in bio-processors
Abstract
Motivated by the idea that one day, probably far in the future, the computers and robots will be architectureless, made of collections of numerous 'intelligent' subsystems or nanomachines able to self-organize each other into computational morphologies with perhaps more computational power than classical electronic-based computers, many studies are burgeoning in different fields (chemistry, biology, condensed matter, quantum physics, ...). Several systems inspired from Nature have indeed been proposed yet for designing unconventional computer architectures using processing modes of various nature and at different scales. The heterogeneous set of natural or artificially designed systems called trail systems, commonly associated to self-driven particles (agents1 ) with tropistic activity (through a communication based on traces let in the environment), is a soft matter with self-organizing properties sufficiently robust and fine for designing biocomputing structures. In this context, individual trails systems could be viewed as single wires and logical gates in a self-organized bio-processor, in the same manner axons are connecting the neural nodes in a neuro-processor. Their efficiency as wires depends on their specific properties which are often related to their scale. The robustness of their self-organization at the microscopic scale level occurring in a noisy environment, can be studied by a model based on effective computing systems (i.e. Turing machines) programmed to behave first as deterministic and perfect trail systems, then as stochastic-working trailing agents subject to randomness.
Fichier principal
Glade-Trail_Systems-Modelling_Complex_Systems_09.pdf (603.41 Ko)
Télécharger le fichier
Origin : Files produced by the author(s)
Loading...