Clemson Intelligent Infrastructure : Enabling Domain Research 'Out in the Wild'

Last update: 12/30/2020

We acknowledge the National Science Foundation (award numbers 1531127 and 1544910), DOJ/NIJ, Cisco, and IBM for partially supporting the research described on this page.

ORIGINAL DIRECTION (circa 2015)

At Clemson University, as well as other research Universities, in the world), there are MANY research activities underway that share a common need for infrastructure to support research 'out in the wild'. Fundamental requirements include the use of devices, platforms, software systems to acquire data from the field, possibly process the data in the field, publish the data to a back-end for further processing as well as for storage and shared access.

Examples of early projects that fall in this direction:

Extreme Orange: Engaging Clemson Students in Systems of Discovery
Intelligent River
SciWiNet - The Science Wireless Network - a project that explored the use of an Mobile Virtual Network Operator (MVNO) for the research community. A project summary is available.

SciWiNet: The Science Wireless Network or SciWiNet is an exploratory NSF project. The project built and trialed a large scale wireless network that supports the academic research community. To accomplish a portion of the project goals, The project had the following three goals:

Requirements gathering : The objective was to learn how the community was currently using public cellular systems for research. We interacted with researchers through trials, surveys, and community workshops and talks.
Trial of an MVNO : Partnered with several cellular companies (Sprint Wireless and Arterra). We decided to use a lightweight form of a Mobile Virtual Network Operator (MVNO). An MVNO resells services from an actual wireless network operator (WNO). Typically an MVNO addresses the needs of a specific user community that might not be addressed by large commercial wireless operators. The research issue was if the academic community would benefit from a wireless operator that focused on its specific needs.
Project Tools and community development: Through trials and surveys, we wanted to identify common or shared infrastructure (including open and configurable Android platforms, measurement software and services) that would benefit the community.

The SciWiNet project web page is located at http://sciwinet.org . We are currently in the third year of this project. The MVNO is no longer in service.

The topic of community-based wireless infrastructure involving public networks to support academic research has been of high interest to NSF for decades. Discussions of a NSF funded MVNO surfaced early in the GENI project planning stages. Our results are based on interactions with hundreds of researchers, students, and faculty across the country. We determined that necessary requirements of the community is access to a diverse and cost effective set of service plans. Further, most research projects will tend to be on and then off such that a large number of active devices are required for a small amount of time. This will become especially true as IoT research becomes more widely accessible to academics. Our requirements were focused on the wireless research community. Our wireless operator partners would not allow us access to internal control or information from the wireless network making the infrastructure of minimal use to wireless networking researchers. As this community is relatively small, it reduces incentives for wireless operators to make interesting networking information or control possible. If we were to consider the needs of the broader academic community (i.e., including NSF programs outside of wireless and networking, perhaps including research funded by other government agencies such as Department of Transportation (DOT), Department of Energy (DOE) and Department of Education), the size of the broad community potentially increases in size by several orders of magnitude. It might be possible to exert more influence on the operators for attention. The recommendation to NSF was that a non-profit organization might be able to better meet the needs of the academic and more broadly the higher education market.

Project output includes:

The needs of academic research community for such capabilities have existed as long as wireless systems where widely available. However, during the 2016-2018 time frame the need for domain research 'out in the wild' exploded. In parallel to the projects listed above, we have had a long-running research agenda in the area of heterogeneous wireless networks (hetnets) and more recently a research direction involving the intersection of advanced wireless technology applied to address problems related to connected vehicles.

Other deployment or measurement oriented research projects:

WiMAX for Public Safety: Leveraging our expertise in DOCSIS systems, we successfully obtained funding from the Department of Justice (DOJ) to provide guidance related to the choice of broadband wireless technology for spectrum that was to be made available to public safety by the FCC (in the 700 MHz region). Our DOJ project page is located at this web link:http://people.cs.clemson.edu/~jmarty/projects/PublicSafety/PublicSafety.html

The research provides an assessment of WiMAX operating at 4.9 GHz with guidance as to how public safety applications might behave in realistic environments. Our results suggested that the 4.9 GHz spectrum is problematic for mobile nodes that have near-line-of-site with the base station. We pointed out that LTE would not address this problem although it would most likely better address interoperability requirements. The results from this work were widely disseminated to the public safety community through participation in tech transfer activities with the public safety community. Several of our publications are listed:

J. Deng, R. Brooks, J. Martin, “Assessing the Effect of WiMAX System Parameter Settings on MAC-level Local DoS Vulnerability”, International Journal of Performability Engineering, 8(2), 806-819 (2012).
J. Westall, J. Martin, “Performance Characteristics of an Operational WiMAX Network”, IEEE Transactions on Mobile Computing, 10(7), 941-953 (2011).
Zhou, Y., Chowdhury, M., Martin, J., Wang, K.C., Westall, J., Kang, X., “Field Performance Study of a Regional WiMAX Network for Intelligent Transportation System Applications”, Journal of Transportation Research Board, Transportation Research Board, (2129), 121-128 (2009).

GENI Projects: As a part of the NSF GENI Spiral 3 grant (GENI WiMAX at Clemson), we deployed a WiMAX network at Clemson involving 3 base stations. Our hetnets work guided the design and implementation of an SDN-based wireless handover technique that is currently being released to the GENI experimentation community. My contribution on GENI was to guide students through various work efforts surrounding the deployment and validation of the WiMAX network at Clemson.

R. Izard, A. Hodges, J. Liu, J. Martin, K. Wang, K. Xu, “An OpenFlowTestbed for the Evaluation of Vertical Handover Decision Algorithms in Heterogeneous Wireless Networks”, Proceedings of 9th International Conference on Testbeds and Research Infrastructures for the Development of Networks & Communities (TRIDENTCOM 2014), (Guangzhou China, May 2014).
K. Xu, R. Izard, K. Wang, J. Martin, “Cloud-based Handoff as a Service for Heterogeneous Vehicular Networks with OpenFlow (short paper)”, Proceedings of the GENI Research and Educational Experiment (GREE2013), (Salt Lake City UT, March 2013).
Hodges, “The Design and Implementation of an Over-the-Top Cloud-Based Vertical Handoff Decision Service For Heterogeneous Wireless Networks”, MS Thesis, August 2014.

Other project output includes: