Epilepsy is a network disease. Identifying the epileptogenic networks from noninvasive recordings is a challenging issue. In this context, M/EEG source connectivity is a promising tool to identify brain networks with high temporal and spatial resolution. In this paper, we analyze the impact of the two main factors that intervene in EEG source connectivity processing: i) the algorithm used to solve the EEG inverse problem and ii) the method used to measure the functional connectivity. We evaluate three inverse solutions algorithms (dSPM, wMNE and cMEM) and three connectivity measures (r(2), h(2) and MI) on data simulated from a combined biophysical/physiological model able to generate realistic interictal epileptic spikes reflected in scalp EEG. The performance criterion used here is the similarity between the network identified by each of the inverse/connectivity combination and the original network used in the source model. Results show that the choice of the combination has a high impact on the identified network. Results suggest also that nonlinear methods (nonlinear correlation coefficient and mutual information) for measuring the connectivity are more efficient than the linear one (the cross correlation coefficient). The dSPM as inverse solution shows the lowest performance compared to cMEM and wMNE.