Node localization in wireless sensor networks has been a topic of active research in recent years. Network connectivity has been shown to be effective with respect to increasing localization performance via redundant nodes. In this paper, we study the connectivity of wireless sensor networks from a network self-localization prospective. To the best of our knowledge, theoretic analyses of the network connectivity for node self-localization have not been previously studied in the networking literature. Assuming a spatial Poisson distribution of network nodes, we modeled the network as random graph and derived an analytical expression that allows the determination of the localizable probability of any an unknown node for given network density /spl rho/. The problems have also been investigated by simulation in various scenarios. Our results are of practical value for designing sensor network considering of node self-localization.