418 | | <br><br><b>Abstract: </b>Computer networking researchers often have access to a few dierent network testbeds (Section 1.2) for their experiments. However, those testbeds are limited in resources; contentions for resources are prominent in those testbeds especially when conference deadline is looming. Moreover, services running on those testbeds are subject to seasonal and daily trac spikes from users all round the world. Hence, demand for resources at the testbeds are high. Some researchers can use other testbeds in conjunction with the ones they are using. Even though each of the testbeds may have dierent infrastructures, and characteristics, in the end, what the researchers receive in return is a set of computing resources, either virtual machines or physical machines. Essentially, those testbeds are providing a similar service, but researchers have to manage the credentials for accessing the testbeds manually, and they have to manually request resources from dierent testbeds in order to setup experiments that span across dierent testbeds. This thesis presents GENICloud, a project that enables the federation of testbeds with clouds. Computing and storage resources can be provisioned to researchers and services running on existing testbeds dynamically from an Eucalyptus cloud. As a part of the GENICloud project, the user proxy (Section 3.4) provides a less arduous method for testbeds administrators to federate with other testbeds; the same serviceiv also manages researchers credentials, so they do not have to acquire resources from each testbed individually. The user proxy provides a single interface for researchers to interact with dierent testbeds and clouds and manage their experiments. Furthermore, GENICloud demonstrates that there are, in fact, quite a few architectural similarities between dierent testbeds and even clouds |
| 418 | <br><br><b>Abstract: </b>Computer networking researchers often have access to a few di |
| 419 | erent network testbeds (Section 1.2) for their experiments. However, those testbeds are limited in resources; contentions for resources are prominent in those testbeds especially when conference deadline is looming. Moreover, services running on those testbeds are subject to seasonal and daily trac spikes from users all round the world. Hence, demand for resources at the testbeds are high. Some researchers can use other testbeds in conjunction with the ones they are using. Even though each of the testbeds may have di |
| 420 | erent infrastructures, and characteristics, in the end, what the researchers receive in return is a set of computing resources, either virtual machines or physical machines. Essentially, those testbeds are providing a similar service, but researchers have to manage the credentials for accessing the testbeds manually, and they have to manually request resources from di |
| 421 | erent testbeds in order to setup experiments that span across di |
| 422 | erent testbeds. This thesis presents GENICloud, a project that enables the federation of testbeds with clouds. Computing and storage resources can be provisioned to researchers and services running on existing testbeds dynamically from an Eucalyptus cloud. As a part of the GENICloud project, the user proxy (Section 3.4) provides a less arduous method for testbeds administrators to federate with other testbeds; the same serviceiv also manages researchers credentials, so they do not have to acquire resources from each testbed individually. The user proxy provides a single interface for researchers to interact with di |
| 423 | erent testbeds and clouds and manage their experiments. Furthermore, GENICloud demonstrates that there are, in fact, quite a few architectural similarities between di |
| 424 | erent testbeds and even clouds |
| 1038 | , "Performance of GENI Cloud Testbeds for Real Time Scientific Application." |
| 1039 | First GENI Research and Educational Experiment Workshop (GREE 2012), Los Angeles, |
| 1040 | 2012. |
| 1041 | |
| 1042 | |
| 1043 | <br><br><b>Abstract: </b>Dedicating high end servers for short-term execution of scientific applications such as weather forecasting wastes resources. Cloud platforms IaaS model seems well suited for applications which are executed on an irregular basis and for short duration. In this paper, we evaluate the performance of research testbed cloud platforms such as GENICloud and ORCA cloud clusters for our real-time scientific application of short-term weather forecasting called Nowcasting. In this paper, we evaluate the network capabilities of these research cloud testbeds for our real-time application of weather forecasting. In addition, we evaluate the computation time of executing Nowcasting on each cloud platform for weather data collected from real weather events. We also evaluate the total time taken to generate and transmit short-term forecast images to end users with live data from our own radar on campus. We also compare the performance of each of these clusters for Nowcasting with commercial cloud services such as Amazon's EC2. The results obtained from our measurement show that cloud testbeds are suitable for real-time application experiments to be carried out on a cloud platform. |
| 1044 | </li> |
| 1045 | <br> |
| 1046 | |
| 1047 | <li> |
| 1048 | <b>Krishnappa, Dilip K. and Lyons, Eric and Irwin, David and Zink, Michael</b> |
1041 | | <li> |
1042 | | <b>Krishnappa, Dilip K. and Lyons, Eric and Irwin, David and Zink, Michael</b> |
1043 | | , "Performance of GENI Cloud Testbeds for Real Time Scientific Application." |
1044 | | First GENI Research and Educational Experiment Workshop (GREE 2012), Los Angeles, |
1045 | | 2012. |
1046 | | |
1047 | | |
1048 | | <br><br><b>Abstract: </b>Dedicating high end servers for short-term execution of scientific applications such as weather forecasting wastes resources. Cloud platforms IaaS model seems well suited for applications which are executed on an irregular basis and for short duration. In this paper, we evaluate the performance of research testbed cloud platforms such as GENICloud and ORCA cloud clusters for our real-time scientific application of short-term weather forecasting called Nowcasting. In this paper, we evaluate the network capabilities of these research cloud testbeds for our real-time application of weather forecasting. In addition, we evaluate the computation time of executing Nowcasting on each cloud platform for weather data collected from real weather events. We also evaluate the total time taken to generate and transmit short-term forecast images to end users with live data from our own radar on campus. We also compare the performance of each of these clusters for Nowcasting with commercial cloud services such as Amazon's EC2. The results obtained from our measurement show that cloud testbeds are suitable for real-time application experiments to be carried out on a cloud platform. |
1049 | | </li> |
1050 | | <br> |
1051 | | |
| 2107 | <b>Fei, Zongming and Yi, Ping and Yang, Jianjun</b> |
| 2108 | , "A Performance Perspective on Choosing between Single Aggregate and Multiple Aggregates for GENI Experime nts." |
| 2109 | EAI Endorsed Transactions on Industrial Networks and Intelligent Systems, |
| 2110 | 2014. |
| 2111 | doi:10.4108/inis.1.1.e5. |
| 2112 | <a href="http://dx.doi.org/10.4108/inis.1.1.e5">http://dx.doi.org/10.4108/inis.1.1.e5</a> |
| 2113 | <br><br><b>Abstract: </b>The Global Environment for Network Innovations (GENI) provides a virtual laboratory for exploring future internets at scale. It consists of many geographically distributed aggregates for providing computing and networking resources for setting up network experiments. A key design question for GENI experimenters is where they should reserve the resources, and in particular whether they should reserve the resources from a single aggregate or from multiple aggregates. This not only depends on the nature of the experiment, but needs a better understanding of underlying GENI networks as well. This paper studies the performance of GENI networks, with a focus on the tradeoff between single aggregate and multiple aggregates in the design of GENI experiments from the performance perspective. The analysis of data collected will shed light on the decision process for designing GENI experiments. |
| 2114 | </li> |
| 2115 | <br> |
| 2116 | |
| 2117 | |
| 2118 | |
| 2119 | <li> |
| 2277 | , "Creating environments for innovation: Designing and implementing advanced experimental network research testbeds based on the Global Lambda Integrated Facility and the StarLight Exchange." |
| 2278 | Computer Networks, |
| 2279 | 2014. |
| 2280 | doi:10.1016/j.bjp.2013.12.024. |
| 2281 | <a href="http://dx.doi.org/10.1016/j.bjp.2013.12.024">http://dx.doi.org/10.1016/j.bjp.2013.12.024</a> |
| 2282 | <br><br><b>Abstract: </b>Large scale national and international experimental research environments are required to advance communication services and supporting network architecture, technology, and infrastructure. Theories and concepts are often explored using simulation and modeling techniques within labs or on small scale testbeds. However, while such testbeds are valuable resources for the research process, these facilities alone cannot provide an appropriate approximation of the real world conditions required to explore ideas at scale. Very large scale global, experimental network research capabilities are required to deeply investigate innovative concepts. For many years, network testbeds were created to address fairly specific, well defined, limited research goals, and they were implemented for fairly short periods. Recently, taking advantage of a number of macro information technology trends, such as virtualization and programmable resources, several network research communities have been developing innovative types of network research environments. Instead of designing traditional network testbeds, research communities are designing large scale, highly flexible distributed platforms that can be used to create many different types of testbeds. Also, rather than creating short term testbeds for limited research objectives, these new environments are being designed as long term persistent resources to support many types of experimental research. This paper describes the motivations for this trend, provides several examples of large scale distributed network research environments based on the Global Lambda Integrated Facility (GLIF) and the StarLight Exchange Facility, including the Global Environment for Network Innovation (GENI), and indicates emerging future trends for these types of environments. |
| 2283 | </li> |
| 2284 | <br> |
| 2285 | |
| 2286 | <li> |
| 2287 | <b>Mambretti, Joe and Chen, Jim and Yeh, Fei</b> |
2267 | | <li> |
2268 | | <b>Mambretti, Joe and Chen, Jim and Yeh, Fei</b> |
2269 | | , "Creating environments for innovation: Designing and implementing advanced experimental network research testbeds based on the Global Lambda Integrated Facility and the StarLight Exchange." |
2270 | | Computer Networks, |
2271 | | 2014. |
2272 | | doi:10.1016/j.bjp.2013.12.024. |
2273 | | <a href="http://dx.doi.org/10.1016/j.bjp.2013.12.024">http://dx.doi.org/10.1016/j.bjp.2013.12.024</a> |
2274 | | <br><br><b>Abstract: </b>Large scale national and international experimental research environments are required to advance communication services and supporting network architecture, technology, and infrastructure. Theories and concepts are often explored using simulation and modeling techniques within labs or on small scale testbeds. However, while such testbeds are valuable resources for the research process, these facilities alone cannot provide an appropriate approximation of the real world conditions required to explore ideas at scale. Very large scale global, experimental network research capabilities are required to deeply investigate innovative concepts. For many years, network testbeds were created to address fairly specific, well defined, limited research goals, and they were implemented for fairly short periods. Recently, taking advantage of a number of macro information technology trends, such as virtualization and programmable resources, several network research communities have been developing innovative types of network research environments. Instead of designing traditional network testbeds, research communities are designing large scale, highly flexible distributed platforms that can be used to create many different types of testbeds. Also, rather than creating short term testbeds for limited research objectives, these new environments are being designed as long term persistent resources to support many types of experimental research. This paper describes the motivations for this trend, provides several examples of large scale distributed network research environments based on the Global Lambda Integrated Facility (GLIF) and the StarLight Exchange Facility, including the Global Environment for Network Innovation (GENI), and indicates emerging future trends for these types of environments. |
2275 | | </li> |
2276 | | <br> |
2277 | | |
| 2695 | </li> |
| 2696 | <br> |
| 2697 | |
| 2698 | |
| 2699 | |
| 2700 | <li> |
| 2701 | <b>Xu, Ke and Wang, Kuang-Ching and Amin, Rahul and Martin, Jim and Izard, Ryan</b> |
| 2702 | , "A Fast Cloud-based Network Selection Scheme Using Coalition Formation Games in Vehicular Networks." |
| 2703 | IEEE Transactions on Vehicular Technology, IEEE, |
| 2704 | 2014. |
| 2705 | doi:10.1109/tvt.2014.2379953. |
| 2706 | <a href="http://dx.doi.org/10.1109/tvt.2014.2379953">http://dx.doi.org/10.1109/tvt.2014.2379953</a> |
| 2707 | <br><br><b>Abstract: </b>Leveraging multiple wireless technologies and radio access networks, vehicles on the move have the potential to get robust connectivity and continuous service. To support the demands of as many vehicles as possible, an efficient and fast network selection scheme is critically important to achieve high performance and efficiency. So far, prior works have primarily focused on design of optimization algorithms and utility functions for either user or network performance. Most such studies do not address the complexities involved in the acquisition of needed information and the execution of algorithms, making them unsuitable for practical implementations in vehicles. This paper proposes a fast, cloud-based network selection scheme for vehicular networks. By leveraging a compute cloud's abundant computing and data storage resources, vehicles can leverage wider scope network information for decision making. Vehicles select best access networks through a coalition formation game approach. A one-iteration fast convergence algorithm is proposed to achieve the final state of coalition structure in the game. Through extensive simulation, the proposed network selection scheme was shown to balance system throughput and fairness with built-in utility division rule of the framework. The algorithm efficiency showed eight-fold enhancement over a conventional coalition formation algorithm. Such features validate the potential of implementation in practice. |