This paper presents a cloud-based and hybrid wireless mesh communication framework for bilevel, nested, distributed optimization of networked clusters of microgrids. The proposed optimization framework implements a diffusion-based, fully distributed algorithm on local wireless network and a quasi-distributed approach on wide-area internet-based cloud. The lower level of the bilevel optimization implements a distributed optimal economic dispatch solution for intramicrogrid among distributed energy resources, and the upper level implements a global optimal dispatch for intermicrogrid energy exchange. To demonstrate industrial applicability of the proposed framework, the IEC 61850 interoperability protocol is adopted to achieve a certain delay performance so that the distributed optimization convergence is guaranteed. First, hardware-based prototype intelligent electronic devices are developed using embedded systems. Then, the bilevel nested optimization algorithm is implemented for integration of networked microgrids. Finally, experimental results are demonstrated from a real-time smart grid testbed featuring realistic microgrids. The results demonstrate that the proposed framework meets communication requirements for distributed optimization of networked microgrids.