Rights statement: ©2019 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.
Accepted author manuscript, 335 KB, PDF document
Available under license: CC BY-NC: Creative Commons Attribution-NonCommercial 4.0 International License
Final published version
Research output: Contribution to Journal/Magazine › Journal article › peer-review
| <mark>Journal publication date</mark> | 1/07/2019 |
|---|---|
| <mark>Journal</mark> | IEEE Pervasive Computing |
| Issue number | 3 |
| Volume | 18 |
| Number of pages | 9 |
| Pages (from-to) | 40-48 |
| Publication Status | Published |
| <mark>Original language</mark> | English |
Proliferation of sensors into everyday environments is resulting in a connected world that generates large volumes of complex data. This data is opening new scientific and commercial investigations in fields such as pollution monitoring and patient health monitoring. Parallel to this development, deep learning has matured into a powerful analytics technique to support these investigations. However, computing and resource requirements of deep learning remain a challenge, often forcing analysis to be carried at remote third-party data centers. In this paper, we describe an alternative computing as a service model where available smart devices opportunistically form micro-data centers that can support deep learning-based investigations of data streams generated by sensors. Our model enables smart homes, smart buildings, smart offices, and other types of smart spaces to become providers of powerful computation as a service, enabling edge analytics, and other applications that require pervasive (in-space) decisioning.