Resource Allocation for Cloud-Assisted Mobile Applications

Abstract

Mobile devices such as netbooks, smart phones, and tablets have made computing ubiquitous. However, such battery powered devices often have limited computing power for the benefit of an extended runtime. Nevertheless, despite the reduced processing power, users expect to perform the same types of operations as they could do using their desktop or laptop computers. We address mobile devices's lack of computing power by leveraging cloud computing resources. We present a middleware that relocates computingintensive parts of Java applications to cloud resources. Consequently, our middleware enables the execution of computing-intensive applications on mobile devices. We present a case study on which we adapt Sunflow, an open-source ray tracing application, to use our middleware and show the results obtained by deploying it on Amazon EC2. We show, via simulations, a cost analysis of using the different resource allocation strategies available on our solution.

Convergence of user and service provider perspectives in mobile cloud computing environment: Taxonomy and challenges

“…Complicating the situation further is the fact that offloading all components may not be beneficial in terms of intended objectives and application partitioning itself exerts the extra overheads on the resource constraint SMDs. On the basis of these tradeoffs, the current computation offloading solutions employ the following types of algorithms for partitioning the application logic: (a) static application partitioning algorithms 32,33 or (b) dynamic application partitioning algorithms 31,34–41 …”

“…However, the undertaken applications may differ in the execution complexity. In case of non‐availability of cloud services, the execution of some applications can be hauled by the SMDs, 32 whereas the execution of other applications is not feasible on the SMDs 33 . So, the discussion papers of this part are further divided into those considering either only the thin client ‐based applications or those taking into account also a possibility of local execution, that is, thick client ‐based applications.…”

“…The main objective of studies focused on thin client‐based application partitioning is to realize the feasibility of highly compute‐intensive applications on the SMDs. In Ferber et al, 33 the SMD indispensably requires the support of cloud surrogates for the execution of Sunflow application. Sunflow application renders three‐dimensional scenes by making use of multicore processors.…”

“…Soyata et al 53 and Hoang et al 77 exploited the on‐demand VM provisioning concept for maximizing profits from the limited resource of cloudlets. Unlike reserved VM provisioning, it also imparts flexibility to the CSPs in terms of increasing/decreasing cloud resources according to the arriving workloads and alternately paves way for improving the cloud resource utilization 23,33,41,78,79 . Besides this, users also need to pay for what they use.…”

A Self-Cloning Agents Based Model for High-Performance Mobile-Cloud Computing

“…Mei et al [9] propose a sharing-aware scheme for latency-tolerant as opposed to real-time mobile-cloud applications, where data sharing across multiple applications is exploited for better outsourcing performance. Ferber et al [10] propose a middleware framework for relocating only computing-intensive parts of Java applications to cloud resources, which does not consider dynamic offloading of other application partitions. Lin et al [11] present a model for energy-aware task scheduling on mobile devices, where tasks are assigned to cores on the device or a cloud resource based on their precedence requirements.…”