The increasingly ubiquitous use of cameras creates an exciting novel opportunity to build camera-based optical wireless networks. Optical wireless is not only a potential low-cost alternative where it can take advantage of existing cameras and light emitting devices, but it's highly directional transmissions can present advantages over radio-frequency (RF) based wireless communications. While the optical channel differs fundamentally from the RF channel, this project recognizes that it also allows multiple spatially separated channels between an array of transmitter elements and the array of camera pixels, akin to an RF multiple-input multiple-output (MIMO) system. This inter-disciplinary project therefore brings together expertise in the areas of mobile networks, communications, and computer vision to analyze, design, and prototype a network stack for such visual MIMO communications. This stack addresses the fundamentally different visual channel and receiver constraints through innovative visual signal acquisition, tracking, interference cancellation, and modulation techniques at the physical layer as well as vision-aware link and MAC layer protocols.Visual MIMO networks can potentially support applications ranging from secure communication between cell phones, over localization of 911 callers through surveillance cameras, to interference-free car-to-car communications. The project also makes an experimental visual MIMO testbed available to the research community at large. In addition to publications, the project takes advantage of WINLAB's biannual industry meetings to disseminate results and provides a variety of appealing educational activities involving K-12 and undergraduate students.
|Effective start/end date||4/1/11 → 3/31/15|
- National Science Foundation (National Science Foundation (NSF))