Opportunistic routing (OR) is a powerful new concept that exploits broadcast nature of wireless medium and spatial diversity of network topology to cope with time-varying wireless links in multihop wireless networks. Its potential to improve the network performance, such as network capacity, has not been well understood and there is also a lack of efficient and distributed routing protocols that fully exploit its advantage to achieve performance optimality at the end-to-end path level.

This project focuses on the theory and protocol design of OR in multihop wireless networks. The results will also be extended to incorporate other emerging wireless technologies, such as multi-rate, multi-channel, etc. There are two main thrusts in this project. The first thrust focuses on a theoretic study on the network capacity and performance bounds achievable by OR. A novel theoretical framework will be established to characterize the wireless interference, multirate capability, time-varying channel fading, and their impact on the performance of OR. A method to compute the throughput bounds between a source and destination pair in a given OR network will be devised. The second thrust is to develop distributed and efficient communication protocols. While existing OR schemes mainly focus on the link layer coordination mechanisms, the emphasis of routing protocol design in this project is on the exploit of the spatial diversity in a larger scale at path level and target to end-to-end performance optimality.

This research will advance the theoretical frontier in characterizing fundamental limits of OR in multihop wireless networks and will provide a better understanding of theoretical guidelines to the efficient OR protocol design in such networks.

Project Report

Simple wireless networks – from the Wi-Fi hotspot at your local Starbucks to cellular network that carries your phone calls – have become commonplace over the past decade. In such networks, signals from a laptop jump to a nearby access point or base station in a single hop. The signals travel the rest of the way over wires. Multi-hop wireless network is a more complex type of wireless networks that is just beginning to find real-world uses. In a multi-hop wireless network, such as a sensor network, an ad hoc networks, or a wireless mesh network, most of the nodes are far beyond the reach of an access point. Signals must hop wirelessly from node to node to reach their ultimate destination. In this NSF CAREER award, Dr. Wenjing Lou focuses on finding the efficient and reliable ways to move data through those web-like wireless networks. Lou is looking into a strategy called "opportunistic routing". The basic idea of opportunistic routing is as follows. Because wireless is a broadcast media, when one node is transmitting, all of the neighboring nodes will receive the packet. Opportunistic routing takes advantage of this fact and tries to make every transmission useful. The decision making about the best path for a packet is made among multiple nodes after the actual transmission having taken place. This approach to routing significantly increases the odds of a successful transmission. While the basic idea behind opportunistic routing is simple, putting it into practice has proven quite difficult and little was known about how to design networks to take full advantage of this technique. With this CAREER award, Lou conducted a comprehensive theoretic study on opportunistic routing that, for the first time, produced a framework that shows the maximum possible throughput that can be achieved by this strategy as well as the optimum way for nodes to share information, to determine the best routing strategy for a data packet, and to coordinate the transmission among neighboring nodes in a distributed fashion. Models and protocols that help network designers to use opportunistic routing effectively have been designed from this framework. Effective opportunistic routing protocols can help create more reliable networks, where information gets through more quickly, with fewer retransmissions. Since multi-hop networks tend to be made up of battery-powered nodes, greater efficiency may well translate to longer battery life, higher network throughput and better reliability. The proposed opportunistic routing strategy was not only applied to the traditional simple channel multi-hop wireless networks, Lou also studied the techniques to use it with a variety of emerging wireless technologies, where a radio can operate on different transmission rates (e.g., multi-rate systems), where radios can operate on different frequency bands (e.g., multi-channel systems), where a node is equipped with multiple radios that allow simultaneous transmissions of multiple signals (e.g., multi-radio systems), and where the opportunistic routing is combined with network coding, another technology to improved wireless network throughput. This project has funded five graduate RAs, four of them have completed their Ph.D. dissertations and one has completed the Master thesis during the course of the project. Two Ph.D. students landed academic positons after the graduation and are currently tenure-track assistant professor in US universities. The project has led to 19 journal publications and 21 conference papers/presentations. Graduate students were supported to attend the conferences and present their research results at the conferences. Additionally, Lou and two graduate students supported by this project authored a book titled "Multihop Wireless Networks: Opportunistic Routing", which presents a comprehensive background to the technical challenges and solutions lying behind opportunistic routing. The book was published by Wiley in 2011. During the course of this project, Lou successfully created a strong wireless networking research program at Worcester Polytechnic Institute (WPI) and then at Virginia Polytechnic Institute and State University (VT). Lou was involved in developed a new undergraduate communications and networking course at WPI. A number of undergraduate students at WPI completed their Major Qualifying Projects (MQPs) under the supervision of Lou. They have benefited through the involvement in the experimental aspects of this research project. Lou also developed two new graduate level courses on wireless security and network security respectively at WPI and VT.

Agency
National Science Foundation (NSF)
Institute
Division of Computer and Network Systems (CNS)
Application #
1156311
Program Officer
Min Song
Project Start
Project End
Budget Start
2011-09-01
Budget End
2014-08-31
Support Year
Fiscal Year
2011
Total Cost
$271,970
Indirect Cost
City
Blacksburg
State
VA
Country
United States
Zip Code
24061