A cloud-assisted crowdsourcing machine-to-machine networking platform for vehicular applications

点击翻译按钮获取中文摘要

Machine-to-Machine (M2M) networks have recently attracted significant attention from both industry and academia. Prominent examples are vehicular communication networks, which enable continuous status monitoring of a massive number of on-road vehicles, as well as smart planning towards an intelligent transport system with minimized traffic congestion and exhaust gas emission. Though vehicles are increasingly capable of both short-range and broadband cellular access, they still have limited communication bandwidth and opportunities. Further, due to their limited Central Processing Unit (CPU) power, they are not able to process data efficiently in an online fashion. Heavy-lifting computation, however, is where cloud computing platforms excel. Large volumes of vehicular data, such as location, speed, acceleration, instantaneous mileage and engine status, can be collected in a crowdsourcing fashion, such that they can be processed in the cloud. This proposed Strategic Project seeks to design and implement an efficient crowdsourcing M2M networking platform with full cloud integration, with a specific focus on the integration of vehicular networks. From a theoretical perspective, one original highlight of this project is to study the combination of coding techniques and storage capabilities of both vehicular and road-side communicating nodes, so that a variety of delay constraints can be accommodated as we maximize information flow rates in vehicular applications. From a practical perspective, our new platform will assist rapid development of a new array of applications based on the results of processing crowdsourced data. The proposed research in this project will be undertaken in close collaboration with two industry partners: Qualcomm Canada a world-renowned innovator in wireless communication technologies; and moj.io, a start-up company specializing in cloud applications for vehicular networks.