The nuclear reactor operation is continuously monitored using on-line surveillance systems, by recording the time-dependent evolution of key signals such as the neutron flux, the coolant temperature, the system pressure, etc. Such recorded data are commonly analyzed using traditional post-processing signal analysis techniques in order to identify potential correlation patterns between the signals and to extract signature characteristics of the examined reactor core. In the present paper, the so-called PSI connectivity analysis methodology is employed for the estimation of the cause-and-effect relationship patterns between simulated reactor’s signals, when stochastic perturbations are introduced using the transient nodal code SIMULATE-3K. The main goal is to assess the capabilities and the limitations of the PSI connectivity analysis methodology for assisting on the spatial localization of the simulated perturbation source in the reactor core, by analyzing and following the information flow paths of the the in-core neutron detector signals. The analysis of the simulated scenarios indicate that the current connectivity analysis could assist on the more precise localization of the noise source when a single detector is modelled in the vicinity of the perturbation, while challenges arise when the perturbation source is detected equally by multiple detectors.