We consider the problem of identifying the source of a rumor in a network. Given a snapshot observation of the network, in which a rumor has been spreading for some time, how to identify the source from which the rumor started to spread? In this paper, we point out the limitations of existing estimators in the literature. As a remedy, we put forth a new estimator by incorporating an independent random observation time. To capture the structure of information flow beyond graphs, our formulations consider rate constraints on the rumor and the multicast capacities for cyclic polylinking networks.