Enabling IoT Analytics across Edge and Cloud Platforms**

点击翻译按钮获取中文摘要

Internet of Things (IoT) generated information has, until now, been largely dispatched to the cloud where data is simply stored, or where computing intensive task are performed for analysis. Current cloud computing frameworks can indeed support processing-intensive applications. But with IoT, not only the explosion in data sent to the cloud causes an explosion in network traffic, but when the time taken to transfer the data is high, critical data may lose their validity, or the decision making can no longer be timely enough. To address such issues, edge computing which explores performing some of the computations on the IoT nodes themselves, or on the nodes through which the traffic is transferred, such as gateways, or base stations, has been promoted. With edge computing, processing of the data is performed closer to the edge, and only digests or exceptions are sent into the cloud for more processing or storage. However, edge computing is still at its infancy and many challenges have to be addressed. In this project we will propose solutions to address core challenges that emanate from pushing processing, specifically analytics, closer to the edge. Namely, we will propose solutions for jointly addressing system complexity such as handling failure, scaling and load balancing; distributing analytics across end-devices, edge and cloud infrastructures to overcome geographical distribution and excessive latency; and addressing security issues when leveraging edge nodes for processing to ensure devices are uncompromised and free from intrusions. The project will be performed with our industrial partner, Trilliant Canada Inc., a leader in purpose-built solutions for industrial IoT where applications require real-time insight and intelligence on data. The developments of this project will enrich Trilliant solutions, which encompass networking, services and software platforms; including UnitySuite, an award winning business intelligence platform. The project will contribute to the training of 8 HQP who will be working on the different research issues addressed by the project, in collaboration with the researchers and our industrial partner.