Memory as a Topological Structure on a Surface Network

Abstract

AbstractA special charged surface network with surface spin half particles on it, that can be arranged in topologically inequivalent ways, is introduced. It is shown that action potential-like signals can be generated in the network in response to local surface deformations of a particular kind. Signals generated in this way carry details of the deformation that create them as a form of plasticity that influences the pathways they traverse leaving a topologically stable helical array of spins: a potential memory substrate. The structure is a non transient alignment of surface spins in response to the transient magnetic field generated by the moving charges present in the action potential-like voltage signals generated since particles with spin have magnetic properties. The structure has a natural excitation frequency that may play a role in memory retrieval. Signal generation and memory storage are proposed to depend on the existence of a surface spin structure. We show that such a surface network can capture the intricate topological features of any connectome in the brain. In addition biophysical properties of such a network are examined in order to constrain predictions of how it may function.HighlightsGlobal method for generating one dimensional non-dissipative voltage signal pulses, analytic expressions for them with numerical examples, memory substrate creation, memory lifetimes, surface distortion waves.