Προηγμένη εφαρμογή εκχώρησης πόρων σε υπολογιστικά νέφη μικρής κλίμακας με τη χρήση κινητών συσκευών
Integrated application for resource allocation in a mobile cloudlet environment
KeywordsMobile cloud computing ; Cloudlet ; ClouSim ; Resource Allocation ; Mobile devices ; Android ; Location Based Services
Nowadays, it is a common practice for mobile devices to offload computationally heavy tasks off to a Cloud. In this way, the processing of intensive mobile device jobs takes place in cloud systems, which have rich computational resources. We call such an environment Mobile Cloud Computing (MCC) architecture. In MCC, mobile devices communicate with remote Clouds using 3G/LTE connections. An alternative MCC infrastructure was revealed later in which mobile devices do not communicate directly with Cloud. It had as a result the implementation of a new MCC model, which is based on Cloudlets. In Cloudlet-Based MCC system mobile devices communicate with nearby Cloudlets using Wi-Fi connections and send to them tasks in order to do the required processing and return the final results back. Cloudlets have greater computational resources than smartphones and tablets, but depending on the workload its resources can be exhausted. In this case, Cloudlet asks for extra resources from Cloud. All Cloudlets are connected to a remote Cloud. This dissertation is focused on an alternative approach of the MCC Cloudlet-Based model, where Cloudlet asks for extra computational resources from mobile devices that are connected to the same Wi-Fi LAN. It presents an experimental method, where smart mobile devices offer computational resources to Cloudlets. In the producing context of the dissertation in question, a methodology for resource allocation in android systems is analyzed. Furthermore, a Cloudlet-Based Cloud infrastructure is simulated and experimental scenarios for delegating tasks to mobile devices are executed.