Background The cortex is thought to process sensory stimuli from the environment by flexible routing of neuronal activity across a hierarchy of functionally specialized neuronal networks. This routing necessitates mechanisms that allow for high fidelity communication of neuronal activity between these networks [1]. It was suggested that synchronization of spiking activity plays a pivotal role in this communication process, based on which two seemingly different mechanisms were proposed. The synfire chain hypothesis postulates the existence of highly organized divergent/convergent connections, which allow the generation and faithful transmission of synchronous spike volleys generated by common drive from presynaptic neurons [2]. By contrast, another model proposes that communication between different brain areas is achieved by creating consistent phase relations between population level oscillations entrained by distinct neuronal networks. These oscillations emerge as a consequence of local interactions between excitatory and inhibitory neurons. So far, synchronization driven by oscillations and synchronization due to a common drive have been considered as dynamical processes of a different nature. Here, we outline a new theoretical framework, which views the appearance of coherent oscillations as a manifestation of common input synchrony spreading along diluted feed-forward networks (FFNs), which, initially, fail to create stable propagation of excitatory spike volleys due to insufficient weight and number of connections. We have tested this working hypothesis by implementing numerical simulations of diluted FFNs. In our network model, each FFN group consisted of recurrently connected leaky integrate-and fire neurons with an excitation-inhibition ratio of 4:1.