Exploiting an elastic 2-tiered cloud architecture for rich mobile applications

This dissertation exploits 2-Tier cloud architecture for rich mobile applications. It is based on the observation that using local resources in close proximity to the user, i.e. local clouds or cloudlets (in the second tier), can increase the quality and performance of mobile applications. In contrast, public cloud offerings (in the first tier) offer scalability at the cost of higher delays, higher power consumption and higher price on the mobile device. We introduce new concept of Space-Time workflow as the modeling framework for mobile applications. Based on this framework we try to define the optimal mobile applications (considering power, price and delay as the Qos). This problem in general is NP-Hard. By considering a class of heuristics called CRAM we try to achieve near optimal solution. We will show the efficiency of our developed techniques through simulation and prototyping of different classes of rich mobile applications such as signal processing and streaming applications.