The CVeT testbed will be composed of about 50 cars, vans and buses of the UCLA campus fleet. Each of these cars will be able to directly connect to the Internet through WiFi access points or, if out of access point coverage, through other cars in WiFi range. This will realize a UCLA campus car Internet backbone. The wired and wireless Internet infrastructure will stretch beyond its boundaries through cars.

Past related projects implement network connectivity on buses, with poor chances of providing low delay data transfers. Furthermore, buses move in fixed patterns and don't show how the network can benefit from random moving cars, just as regular traffic. The CVeT project, by using buses, cars and vans, improves connection chances and aims to both test delay-tolerant and real-time applications on cars.

UCLA provides an ideal "lab" environment to test innovative applications on a significant population set. Vehicular connectivity can be exploited in monitoring, safety, information and entertainment applications. More than a scientific experiment, CVeT will serve UCLA with new applications for its community.

Video Demos

CVeT Demo	Urban Grid	Mobility Comparison
In this demo we show a real experiment on the UCLA campus, where six cars leave a parking lot and exchange data on the move. Just imagine your car is able to discover data you want while moving, you could arrive home and already have all the information and files you need!	We here show an emulation of a terrorist detection system. Police cars, provided with threat-detection sensors (e.g. for threats such as chemicals, radiation, etc.), move around the city and spot a suspicious car. While the suspicious car moves, the police cars track the direction of the threat and keeps the area under control. In case the conventional communication infrastructure is damaged while an attack, the police is still able to communicate car-to-car, spread threat position information and consequently track the terrorist's position	In understanding how to build the future car-to-car network, we need to understand how the car-to-car network will behave when it scales to hundreds, thousands and millions of nodes. Thanks to our collaboration with the Los Alamos National Labs, we are able to simulate communication protocols and algorithms on very accurate vehicular traffic mobility traces.

Acknowledgements

We thank the Los Alamos National Labs, in the person of Dr. Stephan Eidenbenz, for providing us the very detailed mobility traces for Portland, Oregon. We acknowledge the use of iNSpect, from Colorado School of Mines, for producing the video in the Mobility Comparison section.