?Zhenjiang Dong

Improving Performanceof Cloud Computing uting and Big Data Technologiesand Applications tions

?Zhenjiang Dong

Zhenjiang Dong is the deputy head of the Cloud Computing and IT Re?search Institute of ZTE Corporation and a standing member and service team leader of the company’s Com?mittee of Corporate Strategy and Planning Experts.He is also an ex?ecutive director of Chinese Associa?tion for Artificial Intelligence，a pro?fessor of Nanjing University of Sci?ence and Technology，a service computing expert of China Computer Federation.He has been responsible for more than 10 research projects supported by National High?Tech R&D Programs of China("863"programs)，Programs of Core Electronic Devices，High?End Generic Chips，and Basic Software Products of China，and National Science and Technology Major Projectof China.His research inter?ests include cloud computing，big data，andmedia analysis and processing.

C loud computing technology is changing the developmentand usage pat?terns of IT infrastructure and applications.Virtualized and distributed systems aswellas unifiedmanagementand scheduling hasgreatly im?proved computing and storage.Management has become easier，and OAMcosts have been significantly reduced.Cloud desktop technology is develop?ing rapidly.With this technology，userscan flexibly and dynamically use virtualma?chine resources，companies’efficiency of using and allocating resources is greatly improved，and information security is ensured.Inmost existing virtual cloud desk?top solutions，computing and storage are bound together，and data is stored as im?age files.This limits the flexibility and expandability of systems and is insufficient formeeting customers’requirements in differentscenarios.

In this era ofbig data，the annualgrowth rate of the data in social networks，mo?bile communication，e?commerce，and the Internet of Things ismore than 50%. More than 80%of thisdata isunstructured.Therefore，it is imperative to develop an effectivemethod for storing and managing big data and querying and analyzing big data in real time or quasi real.HBase is a distributed data storage system operating in the Hadoop environment.HBase provides a highly expandablemethod and plat?form forbig data storageandmanagement.However，itsupportsonly primary key in?dexing but does not supportnon?primary key indexing.As a result，the data query efficiency of HBase is low and data cannot be queried in real time or quasi real time.For HBase operating in Hadoop，the capability of querying data according to non?primary keys is themost importantand urgent.

The graph data structure is suitable formost big data created in social networks. Graph data ismore complex and difficult to understand than traditional linked?list data or tree data，so quick and easy processing and understanding of graph data is ofgreatsignificanceand hasbecomeahot topic in the industry.

Big data has a high proportion of video and image data butmostof the video and image data is notutilized.Creating valuewith this data hasbeen a research focus in the industry.For example，the traditional face localization and identification tech?nology is a local optimal solution that has a large room for improvement in accura?cy.

This special issue of ZTE Communications embodies the industry’s efforts on performance improvementof cloud computing and big data technologies and appli?cations.We invited four peer?reviewed papers based on projects supported by ZTE Industry?Academic?Research Cooperation Funds.

Hancong Duang etal.propose a diskmapping solution integrated with the virtual desktop technology in“A New VirtualDisk Mapping Method for the Cloud Desktop Storage Client.”The virtualdisk driverhasauser?friendlymode foraccessing desk?top dataand hasa flexible cache spacemanagementmechanism.The file system fil?ter driver intelligently checks I/O requests of upper applications and synchronizesfile access requests to users’cloud storage services.Experi?mental results show that the read?write performance of our vir?tual disk mappingmethod with customizable local cache stor?age isalmostsameas thatof the localhard disk.

“HMIBase:An Hierarchical Indexing System for Storing and Querying Big Data，”by Shengmei Luo et al.，presents the design and implementation of a complete hierarchical indexing and query system called HMIBase.This system efficiently que?ries a value or valueswithin a range according to non?primary key attributes.This system has good expandability.Test re?sultsbased on 10million to 1 billion data recordsshow that re?gardless of whether the number of query results is large or small，HMIBase can respond to cold and hot queries one to four levels faster than standard HBase and five to twenty times faster than theopen?source Hindex system.

In“MBGM:A Graph?Mining Tool Based on MapReduce and BSP，”Zhenjiang Dong et al.propose a MapReduce and BSP?based Graph Mining(MBGM)tool.This tooluses the BSP model?based parallel graph mining algorithm and the MapRe?duce?based extraction?transformation?loading(ETL)algorithm，and an optimized workflow engine for cloud computing is de?signed for the tool.Experiments show that graph mining algo? rithm components，including PageRank，K?means，InDegree Count，and Closeness Centrality，in the MBGMtool has higher performance than the corresponding algorithm components of the BC?PDMand BC?BSP.

Bofei Wang et al.in“Facial Landmark Localization by Gibbs Sampling，”presentan optimized solution of the face lo?calization technology based on key points.Instead of the tradi?tional gradient descent algorithm，this solution uses the Gibbs sampling algorithm，which is easy to converge and can imple?ment the global optimal solution for face localization based on key points.In this way，the local optimal solution is avoided. The posterior probability function used by the Gibbs sampling algorithm comprises the prior probability function and the like?lihood function.The prior probability function is assumed to follow the Gaussian distribution and learn according to fea?tures after dimension reduction.The likelihood function is ob?tained through the local linear SVMalgorithm.The LFW data hasbeen used in the system for tests.The test resultsshow that the accuracy of face localization ishigh.

Iwould like to thank all the authors for their contributions and all the reviewerswho helped improve the quality of the pa?pers.

Call for Papers

Special Issue on Using Artificial Intelligence in Internetof Things Guest Editors:FujiRen,Yu Gu

This special issue of ZTE Communications will be dedicated to development，trends，challenges，and current practices in artificial intelligence for the Internet of Things.Position papers，technology overviews，and case studiesareallwelcome.

Appropriate topics include butare not limited to:

?Information technologies for IoT

?Architectureand Layersof IoT

?AItechnologies for supporting IoT

?Imageand Speech SignalProcessing for IoT

?Affective Computing for IoT

?Information Fusion for IoT

?ArtificialConsciousnessand Integrated Intelligence for IoT

ZTE Communications(http://www.zte.com.cn/magazine/English) is a quarterly peer-reviewed technical journal ISSN(1673-5188) and CODEN(ZCTOAK).It is edited，published and distributed by ZTE Corporation(http://www.zte.com.cn)，a leading global provider of telecommunications equipment and network solutions.The jour?nal focuses on hot topics and cutting edge technologies in the tele?com industry.The journal has been listed in Inspec，the Ulrich’s Periodicals Directory，and Cambridge Scientific Abstracts(CSA). ZTECommunications was founded in 2003 and has a readership of 6000.It is distributed to telecom operators，science and technology research institutes，and colleges and universities inmore than 140 countries.

Finalsubm ission due:Feb.5，2015

Publication date:Jun.1，2015

Please email theguesteditor a brief description of the article you plan to submitby Jan.15，2015.

Subm ission Guideline:

Submission should bemade electronically by email inWORD for?mat.

Guest Editors:

Prof.FujiRen

Univ.of Tokushima，Japan，ren@is.tokushima-u.ac.jp

Prof.Yu Gu

HefeiUniversity of Technology，China，yugu.bruce@gmail.com

of Things has

much attention over the past de?cade.With the rapid increase in the use of smart devices，we are now able to collectbig data on a daily basis.The datawe are gather?ing and related problems are becoming more complex and uncer?tain.Researchers have therefore turned to AIas an efficientway of dealingwith the problems created by big data.