Our perception of the Internet is shifting by two main factors, an explosive growth of proactive mobile devices, and an overabundance of data that is growing beyond tractable operation. Thus, the increasing volume of data is in fact becoming less accessible in terms of coherence and synergy, contrary to what search engines would want us to believe! There was a time when IP address space was the major hindrance. Now, the bottleneck has shifted from connecting new devices to handling their data demands (both generated and requested) and cascading replications over the network. In this talk we overview the growing momentum for Information Centric Networking (ICN); a paradigm that envisions networks built around data, rather than the latter being a mere constituent of stale architectures. We highlight two major factors that drive ICN, namely, globalizing the utility of resources that serve the network architecture and cost-effective data harvesting and delivery. We first elaborate on our research in establishing networks on the fly. As networks grow, and the demand for data readiness becomes more stringent, the adoption of application-specific networks presents a major hindrance. At any given location, we need to find the resources that could collect data, process it, and delivers it to designated backhauls at the least cost. To the first end, we overview our work in real-time data collection over opportunistic and participatory sensing networks, overarching a major realization in VANets. We then present our work in optimizing data delivery in dense and infrastructure-less realizations of transient networks. We consider factors of placement, localization, time delivery constraints and cost of delivery over multi-proprietary networks.
展开▼
