Conference PaperPDF Available

Meta-cloud: A cloud of clouds

Authors:

Abstract and Figures

As smart device hardware and software technologies have advanced, performance of devices has improved and their types and categories have diversified. In addition, as smart devices have become popular, the number of multi-devices held by a single person has increased. Following this trend, data production and management through multi-devices of a single user have increased, and various data sharing services have emerged to meet the need for data sharing across multi-devices. Much attention has been given to data sharing services utilizing a cloud computing environment. Cloud computing is a computing technology to provide IT resources (e.g., software, storage, and server) with internet technology. Data sharing services are provided by virtualizing unlimited storage space and computing resources utilizable via the cloud computing environment. Users have access to unlimited storage space for their devices and can share the same data regardless of which device is used. There are a number of pay or free cloud storage services available, but cloud service providers have their own proprietary interfaces so that users may have to install a specific application program for the use of a cloud or data management service in order to access multiple cloud storage services. In the present paper we define and propose the Meta-Cloud, which is an integrated platform for the use of multiple cloud storage services.
Content may be subject to copyright.
Meta-Cloud : A Cloud of Clouds
Myung-Hoon Jeon, Dong-Joo Choi, Byoung-Dai Lee*, Namgi Kim
Department of Computer Science, Kyonggi University, Suwon, Korea
jmh@kgu.ac.kr, choidongjoo@kgu.ac.kr, blee@kgu.ac.kr, ngkim@kgu.ac.kr
Abstract— As smart device hardware and software technologies
have advanced, performance of devices has improved and their
types and categories have diversified. In addition, as smart
devices have become popular, the number of multi-devices held
by a single person has increased. Following this trend, data
production and management through multi-devices of a single
user have increased, and various data sharing services have
emerged to meet the need for data sharing across multi-devices.
Much attention has been given to data sharing services utilizing a
cloud computing environment. Cloud computing is a computing
technology to provide IT resources (e.g., software, storage, and
server) with internet technology. Data sharing services are
provided by virtualizing unlimited storage space and computing
resources utilizable via the cloud computing environment. Users
have access to unlimited storage space for their devices and can
share the same data regardless of which device is used. There are
a number of pay or free cloud storage services available, but
cloud service providers have their own proprietary interfaces so
that users may have to install a specific application program for
the use of a cloud or data management service in order to access
multiple cloud storage services. In the present paper we define
and propose the Meta-Cloud, which is an integrated platform for
the use of multiple cloud storage services.
KeywordsMulti-devices, Data Sharing, Meta Clouds, Multiple
Cloud
I. INTRODUCTION
As hardware and software technologies of smart devices
have advanced, performance and the diversification of smart
devices have increased. In addition, a wide variety of
application programs are now available to utilize smart device
for various tasks through application programs. For example,
editing, such as documents, photos, and videos, which was
previously possible mainly through desktop computers, can
now be accomplished in various smart devices. In addition,
the popularity of smart devices has resulted in multi-devices
being held by a single person. Therefore, users can create
various data by utilizing their own smart devices, and this has
created a need for efficient data sharing over the multi-device
environment.
Data sharing methods can range from a basic one where
users transfer data manually to a method that uses Bluetooth,
NFC, or Wi-Fi Direct. In particular, cloud storage-based data
sharing services have been developed where data is stored by
means of storage resources of cloud computing and data
sharing via various devices. Thus, users have access to
unlimited storage space for their own devices, and data are
shared by any device through the cloud environment. To
increase competitiveness, cloud service providers not only
provide data sharing services but also provide application
services using the stored data. Therefore, when choosing
cloud storage services for data sharing of multi-devices, users
take service quality into consideration in terms of data sharing
as well as the additional application service categories and
performance. This can result in a number of cloud storage
services being chosen to meet the user’s needs. However,
cloud service providers have their own proprietary interfaces
so that users who want access to multiple cloud storage
services have to install specific application programs.
In this paper we define and propose the Meta-Cloud, which
is an integrated platform for the use of multiple cloud storage
services.
II. RELATED WORK
As the use of multi-devices by a single user has increased
so has the importance and the study of data sharing. A
representative data sharing method between devices is the
device to device (D2D) mode[1][2]. BlueTorrent, which is a
P2P-based file sharing application utilizing Bluetooth, has
been proposed, and performance improvement and feasibility
from simulation results of its contents sharing process have
been validated [3]. In addition, [4] proposed a system that can
reduce cellular network traffic efficiently via D2D
communication between neighboring devices over the 3GPP
LTE cellular network. However, such D2D methods are
dependent on locality between devices and not efficient for
data synchronization. Therefore, a data sharing method
utilizing cloud computing has been highlighted to overcome
such problems. In particular, a data sharing service based on
software as a service (SaaS) [5][6], one of the service models
of cloud computing, provides a data sharing feature along with
an application program service based on stored data. [7]
proposed multimedia cloud computing that provides streaming
and viewer services of media data, such as images, music, and
video, thereby not only providing a storage and sharing
feature but also analyzing data types that can be executed in
smart devices. In addition, [8] proposed a system that can
process complicated queries in portable devices, such as
mobile devices, by providing database service based on
BigData in the cloud storage service environment. As types
and features of such cloud storage-based service have
diversified, users use data sharing services through a number
of cloud storages. Therefore, we propose an integrated
platform that provides data sharing and application services
ISBN 978-89-968650-3-2
891
February 16~19, 2014 ICACT2014
through a number of cloud storage services via a general
interface.
III. PROPOSED PLATFORM
In this section we explain the proposed Meta-Cloud. First,
the overall system structure and components are described;
this is followed by an explanation of the overall service
process flow and internal running flow.
A. System Architecture
Figure 1. Meta-Cloud system architecture
Figure 1 shows the system architecture of the proposed
Meta-Cloud. The system is largely divided into three layers:
Gate, Storage, and Service. The Gate Layer consists of multi-
devices of individual users in which various types and
quantities of devices are present. The Storage Layer consists
of cloud storages that store data; cloud storage provides
individual supplementary services. The Service Layer is a core
layer that performs the Meta-Cloud service, which plays a role
of connecting the Gate Layer and Storage Layer. The function
of the Service Layer is to integrate the information of the
Storage Layer, thereby forming a virtual storage environment
and providing this to the Gate Layer. Users who use the Meta-
Cloud service can share data through the integrated virtual
storage environment provided via the Service Layer without
the need to know the status of the cloud storages in the
Storage Layer. The role of the Service Layer is performed
through the following components:
· Common Interface is an interface, which exists as a
general interface form to use Meta-Cloud to
communicate with smart devices directly in the Gate
Layer,
· Directory Mapper provides users with information of
data shared via Meta-Cloud through the virtual
directory and file structure. It also maps and manages
real and virtual storage information of data.
· Storage Locator is a component that determines cloud
storage for user’s data. Optimum cloud storage is
determined via the Storage Analyzer by considering
cloud storage information and size and types of
resources to be stored.
· Storage Analyzer analyzes support services, storage
space status, and network speed information of cloud
storages interconnected to Meta-Cloud thereby
delivering this to the Storage Locator.
· Cloud Server Connector—if preprocessing is required
before storing data, this can be done by utilizing the
computing resources in the cloud. File encryption is a
typical example of such jobs.
· Access Pattern Analyzer analyzes execution history
information, such as user’s data access, download,
storage, and service request, to summarize the user’s
use pattern thereby discovering intelligence, such as
storage optimization and decisions on data replication
storage.
· Cloud Storage Adaptor is a component that performs a
role of storing data to the cloud storages. This requires
scalability to include another cloud storage that
provides new services.
B. Service Execution Flow
Figure 2. Service Scenario of Meta Cloud
Figure 2 shows the overall service process flow and storage
decision process in Meta-Cloud. The Meta-Cloud service flow
is as follows: Users transfer their data stored in their own
multi-devices to Meta-Cloud to share them. Meta-Cloud
analyses a type and size of transferred data and determines the
optimal storage among the connected cloud storages to store
the data. The most important consideration for such service
flow is that a user is not involved in decisions regarding which
cloud storage stores the data but only transfers data via the
interface of Meta-Cloud to store them. Thus, Meta-Cloud
should determine the optimal storage among the connected
cloud storages. For this, the following three steps are required
for the storage selection process by Meta-Cloud:
ISBN 978-89-968650-3-2
892
February 16~19, 2014 ICACT2014
1) It examines whether storage can be available by checking
each storage status over the cloud storages.
2) It checks whether suitable service for the type of data can
be supported.
3) It checks network status and locality of the storage selected
via the above two steps, thereby determining a storage that has
the optimal transfer performance.
Figure 2 shows an example of the cloud storage
determination process. As shown in Fig. 2, five cloud storages
can be found for the individual user and these are connected
with Meta-Cloud. The inner circle in the cloud storage means
available storage size, while the rectangle indicates available
supplementary services (e.g., editing and viewer feature for
document service). In addition, numbers in the dotted arrow
indicate network speed when cloud storage is accessed via the
interface of Meta-Cloud by a user (for ease of explanation, in
this paper we assumed that a higher number indicates a faster
network speed). Thus, if an individual user tries to share
through Meta-Cloud a 15-MB document file (Z) and a 50-MB
video file (X) created through his/her own smart devices, the
following service flow will occur. First, since the size of
document file Z is 15 MB, Meta-Cloud eliminates Cloud
Storage 4, which has insufficient storage space. Next, service
support for the document type is checked so that Cloud
Storages 3 and 5 are excluded. Finally, Meta-Cloud selects
Cloud Storage 1, which has the optimal network performance
for the current user, to store the document file. Similarly, if a
user tries to store video file X, Meta-Cloud determines Cloud
Storage 5 to store the video file. To summarize, a user can
make use of data sharing services utilizing a number of cloud
storages through only the Meta-Cloud interface without
needing to consider the service flow.
IV. ADAPTIVE DECISION MAKING
In this section, the major functions of Meta-Cloud that
improve efficiency of its performance are explained, and we
describe the decision-making process to effectively conduct
these functions.
A. Availability Guarantee
Figure 3. Data replication Mechanism for availability
Availability guarantee is one of the most important
elements for determining service quality when providing a
data sharing service via the cloud storages. Thus, a guarantee
of efficient availability is required to improve overall service
quality of data sharing.
Basically, availability guarantee in cloud computing-based
data sharing services means providing users with continuous
service. For example, let us assume that a user in the situation
of Fig. 2 stored a document file in Cloud Storage 1. Then, if
an unexpected outage occurred in the Cloud Storage 1 server,
a user cannot access the document file until the server is
recovered. That is, availability cannot be guaranteed. To solve
this problem, we use a data replication mechanism (Fig. 3). As
depicted in Fig. 3, a video file (K) transferred from a user is
uploaded to Cloud Storage 1 via Meta-Cloud. Here, Meta-
Cloud determines Cloud Storages 3 and 4 as suitable
replication cloud storage according to the replication
mechanism so that video file K is stored in both Cloud
Storages 3 and 4. In fact, video file K is stored in Cloud
Storages 1, 3, and 4, which means that this video file is
replicated in Cloud Storages 3 and 4. Thus, if Cloud Storage 1
server experiences an outage so that a user cannot access the
file via the server, Meta-Cloud guarantees the availability of
video file K via Cloud Storage 3 or 4. The key to the data
replication mechanism that guarantees availability is that a
user is not involved with determining service availability of
each cloud storage but only requests the required services
through the Meta-Cloud interface.
B. Storage Optimization
Figure 4. User access pattern based Storage Optimization
The efficient utilization and management of storage is one
of the key elements that improves the data sharing service
quality based on cloud computing. Thus, we provide a storage
optimization feature by utilizing the user access pattern to
improve efficiency of cloud storage management. A
characteristic of the proposed storage optimization feature is
to analyze the recent performing pattern of a user with respect
to storage of high utilization frequency, thereby migrating the
infrequently used data to neighboring cloud storages. Thus, a
user can use cloud storage efficiently, which is utilized
ISBN 978-89-968650-3-2
893
February 16~19, 2014 ICACT2014
frequently. Figure 4 shows an example of storage optimization
in Meta-Cloud. As shown in the figure, user JMH is
performing an upload to share his picture file (A) of 1.5 GB
via Meta-Cloud. In the process of determining cloud storage
for this upload, Cloud Storage 2 is excluded because of lack of
storage space, and picture file A is stored in Cloud Storage 3.
However, in this process, Cloud Storage 2, which was
eliminated earlier, may be considered for storage of user data
due to its insufficient data storage availability, and in turn, it
can be considered as cloud storage that is utilized highly. In
fact, as the execution history of recently performed by JMH is
analyzed in Fig. 4, a high frequency of work performed
through Cloud Storage 2 can be found. Therefore, Cloud
Storage 2, which has insufficient storage space, can be
considered as highly utilized storage by JMH so that analysis
on its utilization frequency of internally stored data should be
required. When the execution log of JMH is examined, data
of work and service types performed through Cloud Storage 2
showed a concentration on pictures. However, regarding the
data storage status of Cloud Storage 2, the largest percentage
of file types is video so that current Cloud Storage 2 is
deemed to be inefficient for data management in this case.
Thus, 30 GB of video file, which has a low execution
frequency among the internal data in Cloud Storage 2, is
migrated to Cloud Storage 1, thereby securing more storage
space for highly utilized Cloud Storage 2. In summary, the
performance process for storage optimization is as follows:
1) Storage Optimization Recognition: This step recognizes
the need of storage optimization, mostly through Storage
Analyzer during the upload process. If Storage Analyzer
recognizes a storage deficiency in the middle of the process of
inspection for available storage of each Cloud Storage while
uploading data of users, it delivers the information of the
corresponding Cloud Storage to Access Pattern Analyzer and
requests an execution history analysis.
2) Execution History Analysis: This step analyzes the access
pattern of a user, which is performed by Access Pattern
Analyzer. During this process, Access Pattern Analyzer
identifies the user access pattern based on the recent execution
history of a user. Here, the following considerations are taken
into account. First, frequency of work performed by a user in
the corresponding Cloud Storage is analyzed. For example, if
a user does not access the Cloud Storage frequently,
performing storage optimization may negatively degrade
efficiency. If frequency of works performed by a user in the
corresponding Cloud Storage is determined as high, Access
Pattern Analyzer analyzes the data and service type performed
in the corresponding Cloud Storage and delivers this to
Storage Analyzer.
3) Storage State Analysis: This step is performed by Storage
Analyzer. Data stored in the corresponding Cloud Storage are
identified and their relationship with information delivered
from Access Pattern Analyzer are analyzed. The critical
elements in this process are to analyze storage distribution of
stored data and utilization frequency, and to discover low-
utilized data by matching information delivered from Access
Pattern Analyzer. If low-utilized data are discovered by the
analysis, data migration process is performed.
4) Data Migration: This step migrates the low-utilized data
in the corresponding Cloud Storage, which is performed by
Storage Locator and Storage Analyzer. In this process, the
decision to migrate cloud storage is performed according to
the optimal storage determination method explained
previously. Once cloud storage to be migrated is determined,
low-utilized data are migrated, and the performance result is
sent to the Directory Mapper.
In summary, the storage optimization process provides
cloud storage selection and data management based on access
patterns of users.
V. CONCLUSION
In this paper we proposed a service platform called Meta-
Cloud by which mutually independent cloud storage services
are integrated to form a virtual storage service environment.
We described the overall system architecture and required
components of the proposed Meta-Cloud as well as the service
execution flow and the features for service quality
improvement. The goal of Meta-Cloud is to provide a general
interface to provide users with a virtual storage service
environment. Thus, users do not need to be concerned about
the storage status of application programs and data in the
cloud storages but share data efficiently utilizing a number of
cloud storages.
The Meta-Cloud platform proposed in this paper is in the
early stages of development so that more research is needed to
improve it. For example, development of its prototype is
needed as well as development of a decision-making
algorithm for various conditions and a definition of the
interface for scalability of the Storage Layer.
ACKNOWLEDGMENT
This work (Grants No. 2013-B04, 2013-0554) was
supported by the Content Convergence Software Research
Center funded by the GRRC program of Gyeonggi Province,
South Korea.
REFERENCES
[1] K. Brett and B. Aazhang. “Cellular networks with an overlaid device to
device network,” IEEE 42nd Asilomar Conference, pp.1537-1541,
2008.
[2] P. Jänis, CH. Yu, K. Doppler, CB. Ribeiro, C. Wijting, K. Hugl and V.
Koivunen, “Device-to-Device Communication Underlaying Cellular
Communications Systems,” IJCNS, pp.169-178, 2009.
[3] (2013) The Bluetooth website. [Online]. Available:
http://www.bluetooth.com/
[4] K. Doppler, M. Rinne, C. Wijting, CB. Ribeiro, and K. Hugl. “Device-
to-device communication as an underlay to LTE-advanced networks,”
IEEE Communications Magazine, p.42-49, 2009.
[5] P. Mell, and T. Grance, "The NIST definition of cloud computing,"
NIST special publication 800.145, 2011.
[6] MP. Papazoglou, "Service-oriented computing: Concepts,
characteristics and directions," Proc. WISE, p.3-12, 2003.
[7] W. Zhu, C. Luo, K. Wang and S. Li, “Multimedia cloud computing,”
IEEE Signal Processing Magazine, p.59-69, 2011.
[8] D. Agrawal, S. Das and A. Abbadi, “Big data and cloud computing:
current state and future opportunities,” The 14th International
Conference on Extending Database Technology, p. 530-533, 2011.
ISBN 978-89-968650-3-2
894
February 16~19, 2014 ICACT2014
Dong-Joo Choi is a senior student at the department of
Kyonggi University, Korea. He is pursuing his B.S
degree. His current research interest includes mobile
cloud computing, open mobile platforms, and mobile
network services. After earning his B.S., he plans to
start M.S. program at Kyonggi University
ISBN 978-89-968650-3-2
895
February 16~19, 2014 ICACT2014
Conference Paper
Web system technology is rapidly used by most of the users and there has been a speedy change in the technology to access the information in past fifteen years and at the same time it is highly recommended to cope up with the existing and emerging technologies. Metadata are integral part of information and at the same time it is very important to enhance metadata. It is highly recommended that efficient information retrieval is the most basic need of any kind of user. Efficient set of information is called the significant document which has efficient storage, effective extraction along with the free from different kinds of error. The study of this research paper is to develop an algorithm for storage of metadata to reduce the complexity of storage algorithm for efficient retrieval of web documents in web system so that user can get efficient retrieval of information in distributed environment. This study aims to provide a framework for efficient collection of metadata in distributed environment to improve the web mining for users. Concerning of knowledge issues on web using different kind of methodologies, this study provides the important information.
Article
Full-text available
In this article we propose to facilitate local peer-to-peer communication by a Device-to-Device (D2D) radio that operates as an underlay network to an IMT-Advanced cellular network. It is expected that local services may utilize mobile peer-to-peer communication instead of central server based communication for rich multimedia services. The main challenge of the underlay radio in a multi-cell environment is to limit the interference to the cellular network while achieving a reasonable link budget for the D2D radio. We propose a novel power control mechanism for D2D connections that share cellular uplink resources. The mechanism limits the maximum D2D transmit power utilizing cellular power control information of the devices in D2D communication. Thereby it enables underlaying D2D communication even in interference-limited networks with full load and without degrading the performance of the cellular network. Secondly, we study a single cell scenario consisting of a device communicating with the base station and two devices that communicate with each other. The results demonstrate that the D2D radio, sharing the same resources as the cellular network, can provide higher capacity (sum rate) compared to pure cellular communication where all the data is transmitted through the base station.
Conference Paper
Full-text available
Service-Oriented Computing (SOC) is the computing paradigm that utilizes services as fundamental elements for developing applications/solutions. To build the service model, SOC relies on the Service Oriented Architecture (SOA), which is a way of reorganizing software applications and infrastructure into a set of interacting services. However, the basic SOA does not address overarching concerns such as management, service orchestration, service transaction management and coordination, security, and other concerns that apply to all components in a services architecture.In this paper we introduce an Extended Service Oriented Architecture that provides separate tiers for composing and coordinating services and for managing services in an openmarketplace by employing grid services.
Conference Paper
Full-text available
Spectrum sharing is a novel opportunistic strategy to improve spectral efficiency of wireless networks. Much of the research to quantify such a gain is done under the premise that the spectrum is being used inefficiently by the primary network. Our main result is that even in a spectrally efficient network, device to device users can exploit the network topology to render gains in additional throughput. The focus will be on providing ad-hoc multihop access to a network for device to device users, that are transparent to the primary wireless cellular network, while sharing the primary network's resources.
Article
Full-text available
In this article device-to-device (D2D) communication underlaying a 3GPP LTE-Advanced cellular network is studied as an enabler of local services with limited interference impact on the primary cellular network. The approach of the study is a tight integration of D2D communication into an LTE-Advanced network. In particular, we propose mechanisms for D2D communication session setup and management involving procedures in the LTE System Architecture Evolution. Moreover, we present numerical results based on system simulations in an interference limited local area scenario. Our results show that D2D communication can increase the total throughput observed in the cell area.
Article
This article introduces the principal concepts of multimedia cloud computing and presents a novel framework. We address multimedia cloud computing from multimedia-aware cloud (media cloud) and cloud-aware multimedia (cloud media) perspectives. First, we present a multimedia-aware cloud, which addresses how a cloud can perform distributed multimedia processing and storage and provide quality of service (QoS) provisioning for multimedia services. To achieve a high QoS for multimedia services, we propose a media-edge cloud (MEC) architecture, in which storage, central processing unit (CPU), and graphics processing unit (GPU) clusters are presented at the edge to provide distributed parallel processing and QoS adaptation for various types of devices.
Conference Paper
Scalable database management systems (DBMS)---both for update intensive application workloads as well as decision support systems for descriptive and deep analytics---are a critical part of the cloud infrastructure and play an important role in ensuring the smooth transition of applications from the traditional enterprise infrastructures to next generation cloud infrastructures. Though scalable data management has been a vision for more than three decades and much research has focussed on large scale data management in traditional enterprise setting, cloud computing brings its own set of novel challenges that must be addressed to ensure the success of data management solutions in the cloud environment. This tutorial presents an organized picture of the challenges faced by application developers and DBMS designers in developing and deploying internet scale applications. Our background study encompasses both classes of systems: (i) for supporting update heavy applications, and (ii) for ad-hoc analytics and decision support. We then focus on providing an in-depth analysis of systems for supporting update intensive web-applications and provide a survey of the state-of-the-art in this domain. We crystallize the design choices made by some successful systems large scale database management systems, analyze the application demands and access patterns, and enumerate the desiderata for a cloud-bound DBMS.
Big data and cloud computing: current state and future opportunities Dong-Joo Choi is a senior student at the department of Kyonggi University, Korea. He is pursuing his B.S degree. His current research interest includes mobile cloud computing, open mobile platforms, and mobile network services
  • D Agrawal
  • S Das
  • A Abbadi
D. Agrawal, S. Das and A. Abbadi, " Big data and cloud computing: current state and future opportunities, " The 14th International Conference on Extending Database Technology, p. 530-533, 2011. Dong-Joo Choi is a senior student at the department of Kyonggi University, Korea. He is pursuing his B.S degree. His current research interest includes mobile cloud computing, open mobile platforms, and mobile network services. After earning his B.S., he plans to start M.S. program at Kyonggi University
The NIST definition of cloud computing NIST special publication 800Service-oriented computing: Concepts, characteristics and directions
  • P Mell
  • T Grance
P. Mell, and T. Grance, "The NIST definition of cloud computing," NIST special publication 800.145, 2011. [6] MP. Papazoglou, "Service-oriented computing: Concepts, characteristics and directions," Proc. WISE, p.3-12, 2003. [7]