Data sharing methods have been studied separately in different types of networks, such as P2Ps, MANETs, and WSNs. Little research, however, has investigated pervasive data sharing across heterogeneous networks due to a lack of infrastructure or testbeds to test these ideas. This project will develop, deploy, execute, and analyze a series of data sharing architectures for P2P systems, MANETs, WSNs, and a federated solution across the heterogeneous networks in GENI. In particular, it will investigate: (1) the performance of individual data sharing systems on GENI, (2) the challenges related to achieving pervasive data sharing across heterogeneous networks, and (3) how the different networks can synergistically leverage each others' advantages. The proposed project creates a unified paradigm to study a variety of research areas. The artifacts and results stemming from this work will serve as a catalyst for future work in different computing domains.
In this project, we have deployed and analyzed data sharing systems with routing algorithms on the GENI testbed including the ORBIT testbed and PlanetLab. We compared the experimental results from the GENI testbed experiments with the results from our simulation and analyzed reasons for the discrepancies. The experiment results demonstrated: 1) conducting repeatable experiments on the testbed is challenging due to wide range of hardware/network variables, 2) comprehensive logging of testbed conditions is as crucial as the experimental data itself to interpret experiment results; a complete experimental logging procedure was developed, 3) experiment results differ from simulation results, and factors contributing to such differences were identified for revisiting design of the P2P system. Same conclusions were arrived with conducting the P2P experiments over PlanetLab. Our project also demonstrated the feasibility of simulating node mobility in a mobile data sharing system on the ORBIT testbed with a fixed node grid in the GENI testbed. We distributed our research results through paper publications. The outcome of this project can contribute to the evolution of the next-generation data sharing system. This project also has educational impact. It was integrated into the undergraduate and graduate course teaching. It created collaborative research opportunities for the students and faculty in Clemson University. Undergraduate students and underrepresented students also participated in this project.
