Grant: Discovery Grants Program - Individual

Future networks must serve human demands in ways never before thought possible. Networks will operate in dynamic environments, interact with users and other networks, respond to malicious attacks and yet, have properties of optimality and self-healing while keeping the user out of convoluted details. Future network architectures will be built on a foundation of autonomy, distribution, and intelligence. To achieve that, we must explore and examine new ways of looking at how we design, observe, and interact with future networks. Modern networks like Vehicular Ad-hoc Networks (VANET) offer a glimpse on future network apprehensions. VANET is a new network initiative vowing to improve transportation safety by applying a range of modern wireless technologies connecting vehicles together and to road infrastructure. VANET initiative aims at saving human lives and pledges economical, environmentally-friendly, and congestion-free highways. To put VANET's size into perspective, it is expected to form a continent-wide network with millions of cars joining, interacting, and leaving the network every second. Current and classical network architectures fail to serve the scale and dynamics of the expected VANET. The immediate goal of this research is to resolve the complex VANET scalability problem. This proposal presents the idea of dividing up VANET into smaller localized manageable clusters by dynamically extracting characteristics from its localized environment. Autonomy, distribution, and intelligence are the unique concepts behind the proposed clustering. This research introduces an autonomous clustering framework that mimics the essence of the localized transportation environments. Clustering VANET would not only play a role in decreasing road fatalities and reducing carbon emissions, it also presents a step towards the grand vision of autonomous intelligent networks.