Mobile Code Offloading from Concept to Practice, and Beyond
Download as PPTX, PDF3 likes1,808 views
The document discusses mobile code offloading, which involves transferring computational tasks from mobile devices to the cloud to enhance performance and responsiveness of mobile applications. It highlights issues such as ineffective code profiling and integration complexities that hinder the practical implementation of code offloading. The conclusion suggests that better characterization of the offloading process and faster execution could improve outcomes, proposing future advancements like energy as a service in offloading systems.
Mobile Code Offloading from Concept to Practice, and Beyond
- 1. Mobile Code Offloading from Concept to Practice, and Beyond Huber Flores huber@ut.ee HKUST, Hong Kong, 2014
- 2. Outline • Background (Concept) • Problem statement (Practice) • Solutions (Beyond) • Conclusions huber@ut.ee
- 3. Background • What is code offloading? – Type of binding between mobile and cloud – Move a computational task from one place to another • What is the purpose? – What, when, where, how? – Release the smartphone from intensive processing – Increase performance of the mobile apps, in terms of responsiveness huber@ut.ee
- 4. Background • Code offloading frameworks – Method – Thread – Class huber@ut.ee
- 5. Background • Which are the components of a code offloading architecture? – Mobile • Code profiler • System profilers • Decision engine – Cloud-based surrogate • Server huber@ut.ee
- 7. What is the problem? huber@ut.ee
- 8. Problem statement • Code offloading outcome is mostly negative – Partial answer to what, when, where, and how huber@ut.ee
- 9. Problem statement • Inaccurate code profiling • Low effectiveness to detect resource intensive code huber@ut.ee
- 10. Problem statement • Scalability of the system huber@ut.ee
- 11. Problem statement • Dynamic configuration Cloud ecosystem huber@ut.ee
- 12. Problem statement • Integration complexity – Code annotations – Introduce the mechanisms within the mobile platform huber@ut.ee
- 13. Can code offloading be utilized in practice to sustain power-hungry applications? huber@ut.ee
- 14. Solutions • Massive data analysis with cloud • Mobile crowdsourcing (implicit) • Exploitation of the cloud surrogate huber@ut.ee
- 15. Evidence-aware Mobile Code Offloading huber@ut.ee
- 16. Evidence-aware Mobile Code Offloading • Characterization of the offloading process huber@ut.ee
- 17. Evidence-aware Mobile Code Offloading huber@ut.ee
- 18. Evidence-aware Mobile Code Offloading huber@ut.ee
- 19. Conclusions and future directions • The characterization of the offloading process can enhance the outcome of code invocation. • Increasing the acceleration of the offloading process is the next reasonable step to overcome a negative impact. • Offloading systems powered by SOA – Energy as a service? huber@ut.ee
- 20. THANK YOU FOR LISTENING… huber@ut.ee