Mobile Data Offloading: An Experimental Evaluation

Abstract

Mobile data traffic is becoming a key issue, causing traffic overload problems for classic resource providers. Tackling this issue, different solutions have been proposed, communed by the goal of alleviating the burden on the infrastructure while guaranteeing service availability. In this context, swarming protocols present the advantage of addressing the problem at its source, e.g., by engaging a lower number of direct connections with the network hosted server(s), while the requesters share these same resources among them. However, these solutions are generally verified only through simulation studies. While simulations consist in a necessary first step, they can be far from the actual outcome of employing real protocols and devices. To this end, we discuss BlueFall, an application platform that allows a rapid development and testing of swarming protocols using off-the-shelf smartphones and PCs. Pursuing our goal, we provide some evidence on two swarming schemes we have implemented in our testbed, profiling different access technologies.