Gustavo de Veciana
Professor and Temple Foundation Teaching Fellow
Wireless Networking and Communications Group (WNCG)
Department of Electrical and Computer Engineering
The University of Texas at Austin

Spatial clustering of users.

Architecture and abstractions for environment and traffic aware system-level coordination of wireless networks. We are exploring a system-level framework to mitigate interference using coarse grained coordination of transmissions across base stations. Our approach is based on collecting and mining measured data capturing a user population's diversity in sensitivity to interference -- see figure to the left. Our research aims at developing abstractions and optimizations which enable such coordination to depend on specific characteristics of the traffic and envronment the system is operating under. Our results to date suggest not only improved performance (capacity), but decreased average power requirements along with substantially more uniform wirless coverage. A key research challenge is that of modelling and optimizing the dynamic coupling arising from interference among wireless devices the same spectrum. Click on picture for more.

Spatial energy contours in sensor net.

Using stochastic geometry to better understand wireless and sensor networks. One of our current reserach thrusts involves developing a better understanding of the spatial character of wireless and sensor networks. For example we are investgating how spatially distributed, end systems, wireless providers, sensors can interfere, compete or cooperate with each other to deliver better service. For example by properly routing information in an ad hoc network of wireless sensor nodes one can significantly reduce the energy expenditures. The figure on the left shows the energy contours associated with different locations for a network of sinks, compressor nodes and sensors. As part of our work we are investigating tradeoffs between compression, routing, and energy conservation. Click on picture for more.

Courtesy Prof. Mauseth.

Building computing systems out of unreliable nanotechnologies. As we scale down devices and interconnects they become increasingly difficult to assemble without defects and susceptible to transient malfunction. I'm working with a team of faculty and students towards devising methodologies and tools to enable future engineers to harness the potential of nanoelectronics. Here is an interesting analogy for one of our research ideas. The figure on the left exhibits a cross-section of a plant. The plant has vessels of different diamaters that it uses to bring up water from the ground. Bigger vessels have higher capacity but are more likely to fail particuarly during dry spells while smaller vessels have lower capacity but are more robust during such spells. Thus, plants use a variety of vessels, so they can tolerate different environmental conditions. Similarly a nanocomputing system can make use of structures with different performance reliability characteristics to achieve better overall performance. Well, we are not exactly trying to copy nature, just use good ideas to engineer computing systems! Click on picture for more.