Inhalt des Dokuments
MA: On-Demand Resource Allocation for Mobile Service Computing
Friday, 11. November 2016
Fog Computing is an upcoming paradigm that aims at overcoming the gap between data acquisition and processing. For instance, recent Internet of Things (IoT) architectures and solutions usually follow a Cloud-centric approach. Data are gathered by sensors and devices at the physical layer and forwarded for processing to data centers at the Cloud layer. This approach has advantages in terms of developing solution for data processing but suffers from increased delay and may lead to network congestion. Fog Computing proposes to move the (pre-)processing of data closer to the sources and utilize devices such as routers or smart phones. Applications domains with non-functional requirements, such as logistics, transportation, eHealth are expected to benefit. However, allocating and sharing resources such as CPU on mobile devices owned by different individuals requires serious effort in terms of isolation or resource and service provisioning.
The objective of this master thesis is to practically design and implement a middleware for on-demand resource allocation, service placement and service provisioning on mobile and embedded devices such as smart phones. In particular, this includes providing a light-weight isolation approach (e.g. OS-level virtualization such as Docker).
Students interested in the topic should have knowledge or relevant Distributed Systems concepts as well as Java programming. The topic is available in German and English language.