Optimization of Device Selection in a Mobile Ad-hoc Cloud based on Composition Score

While tapping onto the mobile device capabilities for execution of resource intensive tasks, it has been proven by many studies that the local device resources are unable to completely perform the task execution. Therefore, it becomes imperative that the intra-device resources are put to productive use by offloading the jobs to a remote location for execution. However, in those settings where the network infrastructure is either expensive or inconvenient to use, the traditional cloud would be beyond reachability. This gave rise to the novel "on-the-fly" forms of computing that enables a computation environment closest to the user like a Mobile Ad-hoc Cloud (MAC). Nevertheless, by following this strategy there are more complex unprecedented problems such as constant device movements, disruptions in the external device to name a few that needs to be addressed. Hence, this paper draws attention to the task scheduling process in an MAC. We propose a linear Programming based model to minimize the number of devices participating in an ad-hoc cloud composition by delineating major constraints. In doing so, it is our endeavor to provide a faster ad-hoc cloud composition formation, which leads to quicker task execution in an MAC.