Cooperative localization plays a key role in location aware service of wireless networks. However, the statistical-based estimator of network localization, e.g., the maximum likelihood estimator (MLE) or the maximum a posterior (MAP) estimator, is commonly non-convex due to the nonlinear measurement function and/or the non-Gaussian disturbance, which complicates the localization of network nodes. In this paper, a novel particleassisted
stochastic search (PASS) algorithm is proposed to find out the optimal solution to the non-convex objective function. Given system statistic information, the proposed PASS method is able to find out the global optimum solution with high probability through the robust search based on particle presentation. Moreover, all network nodes can be localized in a distributed manner to harness the uncertainty of dependent factors involved in localization, such as the network node location uncertainties and spatial correlation of measurements. The associated Cramer-Rao lower bound (CRLB) analysis is also presented to benchmark the distributed PASS-based network localization.