MOBICASE 2014, Prof. Satya: “Cloud-Mobile Convergence is the key of research in mobile computing”

The importance of cloudlets and the convergence of mobile computing and cloud computing are the key topics of our interview with Prof. Mahadev Satyanarayanan (Satya). Prof. Satya is a faculty member in the School of Computer Science at Carnegie Mellon University and an experimental computer scientist who has pioneered research in mobile computing, pervasive computing and distributed systems. Keynote Speaker at the 6th International Conference on Mobile Computing, Applications and Services with his Cloudlets: at the Leading Edge of Cloud-Mobile Convergence, Prof. Satya now gives us his point of view about MOBICASE 2014 and points out his opinion about the latest achievements in mobile computing research

What is your overall impression of MOBICASE 2014? What were the main and, perhaps, the most contested topics and debates that have emerged from the discussions?

satya 2
Prof. Mahadev Satyanarayanan (Satya), Keynote Speaker at MOBICASE 2014

In the time that I was there, it appeared that there was a lot of interest in novel applications and services.   This is consistent with the overall focus and theme of MOBICASE, which is carving out its own distinctive niche relative to other mobile computing conferences such as MobiSys and MobiCom.

What are the latest achievements in mobile computing research? What are the main challenges?

The convergence of mobile computing and cloud computing is the source of some of the most exciting research in mobile computing today. Cloud-based services and computing resources can greatly amplify the intrinsic capabilities of mobile devices.  At the same time, the consolidation of services into a distant cloud data center increases end-to-end latency and thereby makes it difficult to offer crisp interactive response to mobile users. How best to reconcile these opposing forces is the subject of much interesting research today.

Why is it important to analyze cloudlets? Which problems can be solved by architectural solutions offered by cloudlets?

Cloudlets address the problem I mentioned above, namely offering crisp interactive response to mobile users while offering cloud-scale resources and services.  They are essentially small data centers that are located at the edge of the Internet, just one wireless hop away from the mobile devices they are serving. So far, cloudlets have been shown to be valuable in mobile multimedia applications such as face recognition, speech recognition, object recognition, and augmented reality.   They are essential for an emerging class of mobile applications called “wearable cognitive assistance” applications. Cloudlets have also been shown to be valuable in executing edge analytics on high data rate sensors (such as video cameras) in the Internet of Things.  We are continuously expanding the class of mobile applications for which cloudlets are appropriate and effective.