My research focuses on networking support for underwater wireless networks, primarily focused on channel scheduling and medium access protocols. The description below gives a very brief introduction to my research; if you want some additional information please see my publications or contact me.
Underwater networks have several unique characteristics when compared to terrestrial networks due to the use of acoustic communications. Water's conductivity prevents RF signals from covering significant distances at reasonable power levels, so devices communicate acoustically through transducers, which convert electrical signals to pressure waves. Acoustic communication results in a slow propagation speed, about 1500 m/s, a bandwidth dependent on distance, a low data rate, and significant multipath effects from the underwater environment. Additionally, the underwater channel varies more than an RF channel due to water movement and it experiences increased doppler effects from device movement.
The distinctive characteristics of the underwater channel provide many opportunities for researchers. My research attempts to leverage these characteristics in order to provide effective performance in the face of the limitations provided by the network conditions. Most of my work has involved channel scheduling and MAC protocols, but I have also investigated some cross-layer issues by looking and link selection and scheduling together.
Long propagation delays in underwater networks invalidate many techniques used in terrestrial networks, so these methods must be adapted or new schemes developed to provide adequate performance. My work focuses on overlapping the transmissions of multiple devices by scheduling channel access in such a way to prevent collisions and increase the number of packets on the air simultaneously. As part of this work I develop constraints on transmission schedules based on interference patterns within the network and show how nodes can find these schedules using distributed algorithms. I evaluate these algorithms through integer linear programming optimization and simulation.
The distributed algorithms that provide a transmission schedule are adaptations of distance-vector routing protocols. Due to this the channel scheduling algorithms can easily be integrated with a routing protocol. I adapted my previous work to include link selection and link scheduling in the same protocol and investigated its performance through simulation.
Publications related to this work include my paper in GLOBECOM.
I performed an initial investigation into MAC protocols for sensor networks, which resulted in a survey article in the Computer Networks journal. In the article I provide a survey of MAC layer issues in sensor network, a description of proposed protocols at that time, and compare the protocols across various criteria.