Neurocomputing 253 (2017) 135–143
Contents lists available at ScienceDirect
Neurocomputing
journal homepage: www.elsevier.com/locate/neucom
Multi-modal Multimedia Big Data Analyzing Architecture and
Resource Allocation on Cloud Platform
K.P.N. Jayasena
a , b
, Lin Li
a , ∗
, Qing Xie
a
a
School of Computer Science and Technology, Wuhan University of Technology, PR China
b
Department of Computing and Information Systems, Faculty of Applied Sciences, Sabaragamuwa University of Sri Lanka, Sri Lanka
a r t i c l e i n f o
Article history:
Received 15 May 2016
Revised 12 September 2016
Accepted 15 November 2016
Available online 8 March 2017
Keywords:
Multi-modal
Multimedia big data
Resource allocation
Cloud computing
Hadoop & MapReduce
ACO
a b s t r a c t
Multimedia big data analyzing is the new topic that focus on all features of distributed computing sys-
tems that contains of a combination of text, visual and audio modalities. The traditional method to
transcoding multi-modal multimedia big data needs expensive hardware and the amount of data in-
creases transcoding executes a significant burden on the computing infrastructure. Therefore we illustrate
a novel implementation for multimedia big data analyzing and data distribution. Our proposed architec-
ture contains three layers such as service layer, platform layer and infrastructure layer. We design and
implement the platform layer of the system by using a MapReduce framework running on a hadoop
distributed file system (HDFS) and the media processing libraries Xuggler. In this way, our proposed sys-
tem reduces the time for transcoding large amounts of data into specific formats depending on the user
requirements. It provides flexible multimedia record/write interface and we can build large scale multi-
media big data analytic applications based on Hadoop cloud platform. Moreover, we proposed the ant
colony optimization (ACO) algorithm for efficient resource allocation in infrastructure layer. The simula-
tion results demonstrate that the proposed algorithm can optimally allocate VM to achieve a minimal
response time.
©2017 Elsevier B.V. All rights reserved.
1. Introduction
Big data analyzing [1] for basic methods such as classifica-
tion, retrieval, and prediction has become famous for multime-
dia sources such as text, graphics, images, audio, and video. The
MapReduce framework [2] developed by Google is very simple
and flexible for various large-scale data processing functions. This
framework is a powerful, efficient tool to develop scalable parallel
applications to analyze big data on large clusters. While increas-
ingly use large multimedia data, more and more users require the
cloud computing technology. It is essential to handle large data in
an effective way, and to focus on transmission efficiency for mul-
timedia data of different quality [3] . The main target of this pa-
per is to present new architecture of the cloud eEnvironment for
multi-modal multimedia service (CEMMS). The main role of our
CEMMS is to support the improvement of multimedia services that
include audio, image, video and text formats. In this paper, we
present the design and implementation of CEMMS, including the
∗
Corresponding author at: School of Computer Science and Technology, Wuhan
University of Technology, PR China
E-mail addresses: pubudu@appsc.sab.ac.lk , pubudu.nuwanthika@gmail.com
(K.P.N. Jayasena), cathylilin@whut.edu.cn (L. Li), felixxq@whut.edu.cn (Q. Xie).
transcoding function for processing variety of multimedia data in
a parallel and distributed manner. The CEMMS is implemented to
increase the quality and speed of multimedia conversions by in-
tegrating a multimedia conversion module based on Hadoop. It
consists of HDFS for storing large amounts of multimedia data
and MapReduce for distributed parallel processing of these data.
In this way, our CEMMS has a greatest advantage of reducing the
encoding time for transcoding large numbers of multimedia files
into specific formats. This architecture contains three basic lay-
ers: service layer, platform layer and infrastructure layer. The main
involvement of service layer, multi-modal multimedia data collec-
tion layer (MMDL) is to collect the variety of video files created
by media creators such as social network service providers, media
sharing services, and private users, as well as the storage of these
files on the local platform. The cloud distributed & parallel pro-
cessing platform (CDPP) layer is the combination of Hadoop and
MapReduce. The cloud based infrastructure layer for quality of ser-
vice (QoS) based cloud computing is deployed at the infrastructure
layer. The major components for the provision of QoS are load bal-
ancing, resource allocation, network traffic management and secu-
rity. In this paper we presented on the resource allocation mech-
anism in multimedia in cloud platform. We propose a novel ACO
algorithm [4] for the multimedia services in the cloud platform.
http://dx.doi.org/10.1016/j.neucom.2016.11.077
0925-2312/© 2017 Elsevier B.V. All rights reserved.