MCluster: A Software Framework for Portable Device-Based Volunteer Computing

Recent market forecasts predict that the portablecomputing trend will vastly spread, as by 2020 there will bemore than 3 billion LTE device users worldwide. Motivatedby this fact, many companies and research institutes havealready launched research projects that utilize portable devices, voluntarily provided by users, to perform the requiredcomputations. Many such projects employ Berkeley's BOINCmiddleware, since it can support a large variety of stationaryand mobile devices. However, currently available BOINChigh-level APIs, either do not support portable devices orlack advanced processing capabilities (such as inter-node taskdependencies) and/or easiness of use. To resolve these issues, we propose the mCluster software framework for applicationexecution powered by the BOINC middleware on portable devices. mCluster adopts a task-based programming model thatrequires simple, pragma-based annotations of the applicationsoftware, in order to dynamically resolve task dependencies. To evaluate our framework, we have have mapped a scientificapplication from the neuroscience domain on an small-scalednetwork of portable devices. mCluster significantly reducesthe required programming effort and complexity to efficientlymap BOINC-powered applications with task dependencies onportable devices compared to previous approaches.