Introducing numerical bounds to improve event-based neural network simulation

Although the spike-trains in neural networks are mainly constrained by the neural dynamic itself, global temporal constraints (refractoriness, time precision, propagation delays, ..) are also to be taken into account. These constraints are revisited in this paper and their consequences are developed at the simulation level. We first introduce these constraints, review their obvious consequence in terms of the amount of information in a spike train and discusse in details the impact at the level of the network dynamics, thanks to recent theoretical results. The goal of this first part is to provide a comprehensive view of the related technical results. Then, the consequences of these time constraints at the simulation level are developed, showing event-based simulation of time-constrained networks can be impacted and somehow improved in this context: the underlying data-structures are strongly simplified, while event-based and clock-based mechanisms can be easily mixed. These ideas are applied to punctual conductance based generalized integrate and fire neural network simulation, while spike-response model simulation is also revisited within this framework. These mechanisms are compared using existing benchmarks.