Oscillations in a Fully Connected Network of Leaky Integrate-and-Fire Neurons with a Poisson Spiking Mechanism

Understanding the mechanisms that lead to oscillatory activity in the brain is an ongoing challenge in computational neuroscience. Here, we address this issue by considering a network of excitatory neurons with Poisson spiking mechanism. In the mean-field formalism, the network's dynamics can be successfully rendered by a nonlinear dynamical system. The stationary state of the system is computed and a perturbation analysis is performed to obtain an analytical characterization for the occurrence of instabilities. Taking into account two parameters of the neural network, namely synaptic coupling and synaptic delay, we obtain numerically the bifurcation line separating the non-oscillatory from the oscillatory regime. Moreover, our approach can be adapted to incorporate multiple interacting populations.