PaperlayerView exact revision on arXiv

Source-linked AI summary

Research On Mobile Cloud Computing: Review, Trend, And Perspectives

Han Qi, Abdullah Gani

arXiv:1206.1118v1cs.DC

TL;DR

Mobile cloud computing combines mobile and cloud computing, but its development is constrained by mobile-device and wireless-network limitations. This paper reviews MCC’s background, principles, characteristics, research projects, challenges, and future directions, concluding with optimization approaches centered on devices, communication, and application-service division.

  • Problem

    MCC requires seamless integration of cloud advantages with mobile devices despite limited processing capacity, storage, battery life, communication quality, and unclear task-division strategies.

  • Method

    The paper reviews MCC’s background and principles, characteristics, infrastructure, recent research projects, challenges, and future research directions.

  • Results

    The paper identifies three main MCC optimization approaches: addressing mobile-device limitations, improving communication quality, and dividing application services.

  • Takeaways & Limitations

    Virtualization, image technology, task migration, and selective application division are presented as approaches for improving mobile-cloud performance and resource use.

  • Takeaways & Limitations

    Resource and performance disparities between mobile devices and other computing systems remain a major MCC challenge.

Abstract

from arXiv · show

Mobile Cloud Computing (MCC) which combines mobile computing and cloud computing, has become one of the industry buzz words and a major discussion thread in the IT world since 2009. As MCC is still at the early stage of development, it is necessary to grasp a thorough understanding of the technology in order to point out the direction of future research. With the latter aim, this paper presents a review on the background and principle of MCC, characteristics, recent research work, and future research trends. A brief account on the background of MCC: from mobile computing to cloud computing is presented and then followed with a discussion on characteristics and recent research work. It then analyses the features and infrastructure of mobile cloud computing. The rest of the paper analyses the challenges of mobile cloud computing, summary of some research projects related to this area, and points out promising future research directions.

I. INTRODUCTION

Cloud computing centralizes computing and services, while mobile cloud computing combines this model with ubiquitous mobile networks to serve mobile devices. The paper frames MCC as a developing paradigm whose benefits must be balanced against mobile-device and wireless-network constraints.

  • Cloud computing provides Internet-based services from clustered systems that centralize computing, services, and applications.
  • Mobile Cloud Computing combines mobile computing and cloud computing through ubiquitous mobile networks and numerous access terminals.
  • MCC virtualizes and distributes resources across computers, making them available to smartphones and other portable terminals.
  • MCC must improve mobile devices by exploiting cloud advantages while addressing limited device resources and computing ability for efficient access.

II. BACKGROUND

Mobile computing links mobile hardware, software, and communication infrastructure so users can interact while moving. Its defining operating conditions include heterogeneous connectivity, interruptions, asymmetric communication, and limited reliability.

  • A. Mobile Computing: Mobile computing comprises mobile hardware, mobile applications, and communication infrastructure designed to remain transparent to end users.
  • A. Mobile Computing: Mobility allows mobile nodes to connect with other nodes, including fixed wired-network nodes, while moving.
  • A. Mobile Computing: Mobile nodes operate across diverse network conditions, ranging from high-bandwidth wired networks to low-bandwidth wireless networks or disconnection.
  • A. Mobile Computing: Battery, communication, and network constraints cause mobile nodes to disconnect and reconnect with wireless networks.
  • A. Mobile Computing: Mobile networks have asymmetric communication because servers and access points provide stronger send/receive capability than mobile nodes.
  • A. Mobile Computing: Signal interference and snooping reduce reliability, requiring security consideration across terminals, networks, databases, and applications.

2) Challenges:

The paper presents cloud computing as a response to growing software and hardware demands, while emphasizing that its definitions remain non-consensual. It characterizes cloud computing as a virtualized, large-scale Internet service paradigm.

  • 2) Challenges:: Mobile computing networks face signal disturbance, security, hand-off delay, limited power, low computing ability, and weather- or building-sensitive QoS.
  • B. Cloud Computing: Cloud computing emerged as users needed faster CPUs, larger storage, and higher-performance operating systems to keep pace with software development.
  • B. Cloud Computing: Cloud computing lacks a consensual definition, with accounts variously emphasizing cloud storage, cached client access, or virtual-machine-based distributed computing.
  • B. Cloud Computing: Cloud computing draws on established areas including distributed computing, grid computing, service-oriented architectures, and virtualization.
  • B. Cloud Computing: The paper treats cloud computing as a large-scale economic and business paradigm centered on virtualization and Internet-based services.

1) Framework:

The cloud-computing framework is organized into infrastructure, platform, and application layers. These layers expose virtualized computing, storage, development environments, and end-user software as services.

  • 1) Framework:: The cloud-computing framework consists of infrastructure, platform, and application layers.
  • Infrastructure layer: The infrastructure layer virtualizes physical servers and storage into resource pools that provide computing, storage, and network services as IaaS.
  • Platform layer: The platform layer supplies parallel-programming environments, distributed storage and file systems, and management tools for developers.
  • Application layer: The application layer delivers software, applications, and customer interfaces to end users as SaaS through clients or browsers.
  • Cloud features: Virtualization lets users access cloud resources through browsers without maintaining their own data centers, while supporting virtual-machine load migration.
  • Cloud features: Cloud systems improve reliability by transferring and backing up data on other machines when servers or virtual machines fail.
  • Cloud features: Large-scale cloud systems may contain thousands of servers and PCs to provide supercomputing capability and mass storage.
  • Cloud features: Autonomous cloud systems configure and allocate hardware, software, and storage resources on demand without exposing management to end users.

3) Challenges:

Cloud computing still requires secure, efficient, monitored, and standardized services, while mobile devices add hardware and sensing capabilities that shape mobile cloud computing.

  • Cloud computing needs safer and more efficient service mechanisms because it relies on numerous third-party software components and infrastructures.
  • Cloud providers must schedule resources efficiently to reduce the electricity consumed by data centres.
  • Service performance must be monitored because cloud computing uses Service Level Agreements between users and providers.
  • Uniform application interfaces are needed to make cloud services simple and convenient for providers.
  • Modern smartphones combine mobile applications with sensing modules for navigation, optics, gravity, and orientation.

A. Concept and principle

Mobile cloud computing combines mobile devices with cloud services by moving intensive computation and storage to distributed cloud resources. Its architecture reduces device-side requirements, but remains constrained by mobile resources, connectivity, and service quality.

  • A. Concept and principle: Mobile cloud computing provides cloud-based services through the Internet and mobile devices by combining mobile computing with cloud computing.
  • A. Concept and principle: Intensive computing, data storage, and mass information processing are transferred to the cloud, reducing mobile devices’ required computing capability and resources.
  • A. Concept and principle: Mobile devices connect to cloud computing through hotspots or base stations using 3G, WiFi, or GPRS, while major computation and data processing occur in the cloud.
  • A. Concept and principle: Monitoring and calculating functions are used to maintain quality of service until a requested connection is completed.
  • 1) Limitations of mobile devices:: Limited computing capability and energy resources restrict smartphones’ ability to deploy complicated applications.
  • 1) Limitations of mobile devices:: Although processing capacity, storage, battery time, and communication improve, major variations in these capabilities remain a challenge for mobile cloud computing.

2) Quality of communication:

Mobile cloud computing must accommodate changing wireless conditions, latency, mobility, and device limitations. Proposed approaches divide processing between mobile devices and cloud resources, including virtualization, edge deployment, and adaptive offloading.

  • 2) Quality of communication:: Mobile cloud connections have changing data-transfer rates and discontinuous connectivity, while mobile users can experience higher network latency and handover delays.
  • 3) Division of application services:: Compute-intensive and data-intensive applications may be unsuitable for mobile devices, so core computing tasks can be processed in the cloud while devices handle simpler tasks.
  • 3) Division of application services:: Quality-guaranteed service design considers application division, code-offload latency, bandwidth, user-oriented performance, adaptation, and resource consumption.
  • 2) Quality of communication:: Regional data centres and edge processing nodes are proposed to improve wireless bandwidth suitability and reduce data-delivery time.
  • 3) Division of application services:: Virtualization and image technologies can duplicate mobile devices in the cloud for data-intensive and energy-intensive computing.
  • 3) Division of application services:: CloneCloud reduced face-tracking execution time from almost 100 seconds on a smartphone to 1 second in the CloneCloud environment.
  • 3) Division of application services:: CloneCloud can reduce battery consumption, but handover delay, bandwidth limits, and signal blind zones can make it unavailable.
  • 3) Division of application services:: Virtualized Screen selectively offloads screen rendering and energy-intensive applications while keeping lightweight operations on smartphones to reduce interaction delay and power consumption.

2) Elastic Applications:

Elastic application approaches distribute application components between mobile devices and cloud resources to extend device capabilities and improve performance. The reviewed systems use dynamic partitioning, middleware-based deployment, and configurable execution across mobile and cloud environments.

  • Researchers developed and extended application-partitioning algorithms to provide more effective mobile cloud applications.
  • CloneCloud-based algorithms dynamically migrate application partitions to remote cloud servers.
  • AlfredO models applications as consumption graphs and automatically selects optimal modules for smartphone–cloud distribution.Its architecture includes AlfredOClient and Renderer on the client, with AlfredOCore on the server.
  • AlfredO computes deployment plans, sends descriptors and service lists to clients, and remotely executes application parts to save battery energy and extend resources.
  • The elastic application model partitions applications into Weblets and dynamically determines deployment, cloud-resource allocation, and network selection.

3) Migration Optimization:

Migration optimization addresses mobility-related transmission and service continuity while mitigating communication delay and mobile-resource constraints. Cloudlet and Hyrax illustrate complementary infrastructure approaches, but Hyrax measurements expose substantial performance limitations on smartphones.

  • Optimal migration mechanisms can reduce interaction delay, enhance processing capability, and improve user experience during mobile data transmission or service delivery.
  • Cloudlet uses virtual-machine technology to provide rapidly customized services near mobile devices and address bandwidth-induced delay between devices and cloud.The approach targets long WAN latency and differing access bandwidths.
  • Hyrax creates a mobile cloud platform from Android phones by adapting Hadoop to run across smartphone nodes.
  • Hyrax deploys smartphones as Hadoop slaves while keeping the master on a PC, with NameNode and JobTracker as background services and HDFS for storage.
  • 15 times as long as PC in Map and Reduce procedures was required by the tested smartphones.The evaluation used 10 Android G1 phones and 5 HTC Magic phones across five workloads.

IV. OPEN RESEARCH ISSUES

Although mobile cloud computing projects have been deployed worldwide, business implementation remains incomplete and further research is needed across several aspects.

  • Mobile cloud computing still faces a substantial gap between existing research projects and broad business implementation.

A. Data delivery

Data delivery remains a central mobile cloud challenge because mobile devices have constrained resources and wireless communication can limit access and performance. The paper emphasizes task partitioning, reduced wireless transfer, virtualization, and elastic service division as optimization approaches.

  • Resource constraints create challenges for cloud accessing, consistent accessing, and data transmission in mobile devices.
  • Task division can offload sub-tasks to the cloud, but the paper identifies the lack of an optimal device-versus-cloud allocation strategy.
  • Different features of mobile devices and PCs prevent direct transplantation of PC services, motivating mobile-specific interactive services.
  • The paper identifies virtualization, image technology, task migration, reduced wireless data delivery, and elastic application division as key MCC optimization approaches.It notes that bandwidth upgrades can improve performance but impose additional user cost.
Loading 1206.1118v1…