Abstract

. This paper deals with the propagation velocity of synfire chain activation in locally connected networks of artificial spiking neurons. Analytical expressions for the propagation speed are derived taking into account form and range of local connectivity, explicitly modelled synaptic potentials, transmission delays and axonal conduction velocities. Wave velocities particularly depend on the level of external input to the network indicating that synfire chain propagation in real networks should also be controllable by appropriate inputs. The results are numerically tested for a network consisting of `integrate-and-fire' neurons. 1 Introduction The concept of synfire chains has been introduced by Abeles [1] in order to explain precisely correlated spike events observable in the cortex on surprisingly large time-scales of up to hundreds of milliseconds. The main idea is that highly specific spatio-temporal firing patterns occur in the brain in such a way that synchronously firing pools ...

Citations