Projects Involved

I have been involved in various academic and engineering projects. I have compiled the list below to give you a feeling as to what kind of projects I have been into. Feel free to browse my research interests and publications for other details you might be after. Also, this recent article by NetworkWorld on WINLAB might be a good point to learn more about our vision at WINLAB. ORBIT ORBIT (Open Access Research Testbed for Next-Generation Wireless Networks) is a two-tier laboratory emulator/field trial network testbed that supports evaluation of protocols and applications in real-world settings. ORBIT is funded by major NSF grants and was born out of a collaborative effort between several university research groups in the NY/NJ region: Rutgers, Columbia, and Princeton, along with industrial partners Alcatel-Lucent Bell Labs, IBM Research and Thomson. ORBIT is being developed and operated by WINLAB, Rutgers University. I have been a member of the core development team since 2004 and worked on various design tasks. I have also used ORBIT testbed in my research for high density wireless LANs as well as vehicular networking. For a light introduction, take a look at this MIT Technology Review article about ORBIT, covering ORBIT and Pervasive Wireless as one of the ten most significant emerging technologies. Our team has recently been awarded the fourth annual Alexander Schwarzkopf Prize for Technological Innovation from the I/UCRC Association, a voluntary, independent organization of past and present members of the NSF’s Industry/University Cooperative Research Center program. LOCATION STACK FOR FUTURE INTERNET This project is funded by FIND (Future Internet Design) initiative of the NSF NeTS research program. As the Internet begins to encompass a larger and more mobile set of devices, including our cars, portable phones, and sensing nodes, a major challenge lies in the integration of location information into the network architecture. There are several key challenges in defining such a location-aware architecture. First, the architecture should include a distributed location service that can track and answer queries about the current positions of billions of moving objects. Because of frequent position changes the load on this service will be higher than on the current domain name system. Second, the architecture must cope with translations between different representations of location and accuracy information. Third, the architecture must be suitable for hierarchical network structures and allow routing between wired and wireless network components. Examination of cross-layer designs that support the use of location information at different (traditional) network layers is one of the key objectives of this project. I have been participating in this project as a research assistant, investigating the kinds of support we need from future Internet for better wireless experience. MULTI-RADIO COEXISTENCE I was involved in this project while interning during summers of 2005 and 2006 at Intel Corp. Communications Technology Lab.  Even today, small laptop platforms with six or more digital communication radios (such as WLAN, WiMAX, GSM, UWB, GPS, Bluetooth, DTV) are not uncommon. Sharing the same spectrum resources or not, concurrent use of these radios create varying degrees of adverse effects on each others performance, some noticeable by the end-user. Objective of the project was first to evaluate and measure common coexistence scenarios using real platforms under typical workloads to find out degree of problem severity. Then several solutions, including a software scheduling for coexistence coordination, have been investigated for orchestrating multiple-radios on our future wireless-capable devices. VEHICULAR NETWORKING In this project, we investigate various forms of vehicle-to-infrastructure (V2I) and vehicle-to-vehicle (V2V) communications to understand the potential gains from optimizations such as fast signaling, diversity, directionality and location awareness. We have been developing a vehicular networking testbed as an extension to ORBIT testbed outdoor infrastructure. This extension is composed of a number of vehicular mobile nodes and fixed-location nodes with proper outdoor antennas to provide us with an opportunity to test our designs in a real-world environment. Our setup currently has one trunk-installed ORBIT outdoor node in a passenger car and five modular seat-tray-installed outdoor nodes. Fixed-location nodes are operational at five different places with ten nodes in ORBIT facility at NJ Tech Centre, North Brunswick and three locations in Rutgers University Busch Campus at Piscataway.   HIGH DENSITY WIRELESS LANs Wireless LANs are being deployed at an unprecedented scale. Hundreds of laptops in a conference room, trying to hunt down a robust connection to the Internet, is not an unreal scenario anymore. However, we are still far from understanding scalability properties of IEEE 802.11 wireless LANs, especially under typical real-world usage scenarios. This project is an effort towards understanding the limits of high density wireless LANs, beyond the overly simplistic 'more clients have to mean degraded system performance' point of view. We have been carrying out ORBIT testbed experiments with up to 350 wireless stations and up to 15 access points to break down the effects of arrival patterns, MAC parameters, transport protocols, workload type etc.   AVAILABLE BANDWIDTH ESTIMATION This project is about practical methods that allow for accurate available wireless bandwidth estimation. An accurate method should account for neighboring node interference (even when perceived in the form of undecodable packets) and flow self-interference. We have investigated applications of packet-probing method of wired networks to wireless networks for this purpose. Our prototype system on ORBIT testbed showed promising results for better wireless resource management that is key to success for multimedia applications in wireless mesh and ad hoc networks.