Stochastic modeling and analysis of opportunistic computing in intermittent mobile cloud
Deze Zeng, Song Guo, Ivan Stojmenović, Shui Yu · 2013
We have witnessed a rapid evolvement in the embedded mobile devices, such as mobile phones and vehicles, with various processing capabilities, some of which are even beyond the common computers. They constitute an unstructured distributed environment with huge computation potential. On the other hand, intermittently connected networks provide a way for delay-tolerant communication between these devices. Such evolution enables a new opportunistic delay-tolerant mobile computing paradigm, called Intermittent Mobile Cloud (IMC), where all transmissions are made over an intermittently connected network and the computation tasks are off oaded to certain service providers that voluntarily share their computational resources in the cloud. In this paper, we first study the capacity of epidemic routing using Random Linear Network Coding (RLNC), which is a critical and fundamental communication issue in IMC. The results are then applied to provide further insights towards a design of efficient bandwidth allocation scheme to minimize the expected completion time of a suite of parallel tasks for a service requester in IMC.