Agile Data Streaming for Grid Applications

Size: px
Start display at page:

Download "Agile Data Streaming for Grid Applications"

Transcription

1 Agle Data Streamng for Grd Applcatons Wen Zhang, Junwe Cao 2,3*, Ysheng Zhong,3, Lanchen Lu,3, and Cheng Wu,3 Department of Automaton, Tsnghua Unversty, Bejng 00084, Chna 2 Research Insttute of Informaton Technology, Tsnghua Unversty, Bejng 00084, Chna 3 Tsnghua Natonal Laboratory for Informaton Scence and Technology, Bejng 00084, Chna * Correspondng emal: jcao@tsnghua.edu.cn Abstract The grd s proposed to be the new computng nfrastructure for cross-doman resource sharng and collaboraton n vrtual organzatons. Whle most current work n grd computng communty focus on adaptablty of grd resource management, aglty s proposed n ths work as another essental requrement, amng to provde mass customzaton and personalzaton on grd computng servce provson. Ths work s focused on data streamng aspects of grd aglty. Snce volumes of data streams are usually extremely hgh but avalable bandwdth and storage are often very lmt n a grd envronment, an enablng envronment for grd data streamng applcatons s mplemented wth supports of on-demand data transfers and just-n-tme data cleanups n ths work. The mplementaton ncludes performance sensors, predctors and a scheduler, wth real-tme measurements, predcton and schedulng capabltes. The GrdFTP s utlzed and data transfer schedules are performed usng the on-the-fly adjustable GrdFTP parallelsm. Expermental results show that the data streamng envronment s agle enough to meet dynamcally changng applcaton data processng requrements and scale well regardng to storage and bandwdth usage.. Introducton Aglty s consdered as an essental feature to many complex system engneerng, e.g. busness [], manufacturng [6] and software development [9]. Aglty represents flexblty of response to dynamcally changng users requrements. Aglty provdes mass customzaton and personalzaton of servces wth extreme flexblty wthout loosng effcency. Ths work s focused on aglty for grd data streamng applcatons. Data management s one of the most challengng ssues n grd mplementaton. Whle most exstng research on data grds prefer to a brng-program-todata approach, grd applcatons such as astronomcal observatons, large-scale smulaton and sensor networks may requre brng-data-to-program supports. For example, LIGO (Laser Interferometer Gravtatonal-wave Observatory) [6,3] s generatng TB scentfc data per day and tryng to beneft from processng capabltes provded by the Open Scence Grd (OSG) [20]. Snce most OSG stes are CPU-rch and storage-constrant wth no LIGO data avalable, data streamng supports are requred n order to utlze OSG CPU resources. In a data streamng scenaro, data are transferred and processed constantly as f data were always avalable from local storage. Meanwhle, processed data have to be cleaned up to save space for the subsequently comng data. In ths work, a data streamng envronment s mplemented wth supports of on-demand data transfers and just-n-tme data cleanups. The mplementaton ncludes several modules, responsble for real-tme measurements, performance predcton, and resource schedulng, respectvely. The envronment s ntegrated wth Condor [9] for job allocaton and Globus [4] for data transfers usng GrdFTP [3]. Data transfer schedules are performed usng the on-the-fly adjustable GrdFTP parallelsm. Expermental results show that aglty of our data streamng envronment can meet dynamcally changng applcaton data processng requrements wth good scalablty. Some exstng work on data streamng management are derved from database management systems, e.g. STREAM [4], Aurora [7], NagaraCQ [8], StatStream [22], and Ggascope [0]. Most of data streamng research n context of grd computng [5,5,7,8] are applcaton specfc and schedulng ssues are not addressed. The followng software mplementaton s most smlar to the work descrbed n ths paper: Pegasus [2,2]. Pegasus handles data transfers, job processng and data cleanups n a workflow manner. Data transfers and processng are executed sequentally. In our envronment, data

2 streamng s performed smultaneously as jobs are processed. By carefully optmzng data streamng, our envronment makes requred data avalable n an on-demand and just-n-tme manner. Streamlne [,2]. Streamlne schedules streamng applcatons on HPC resources usng heurstc methods. Streamlne s bascally stll a job scheduler wth consderaton of data streams; our envronment s focused on schedulng data streamng nstead of job processng wth consderaton of system aglty. Ths paper s organzed as follows: Secton 2 provdes an overall ntroducton to data streamng for grd applcatons, whch s mplemented n detals n Secton 3, usng repertory polces, real-tme measurements, performance predcton and schedulng technques. Expermental results are ncluded n Secton 4 and the paper concludes n Secton Grd Data Streamng A detaled ntroducton to grd requrements for data streamng s provded n ths secton. Gven an example of grd applcatons, two major data management methods are proposed: on-demand data transfers and just-n-tme data cleanups, whch are mplemented n Secton Grd requrements Tradtonal grd computng applcatons prefer to a brng-program-to-data approach. Snce the cost for transferrng large mount of data s always hgh, t s natural that users try to submt ther jobs to computers wth data already avalable and only return data analyss results nstead of transferrng raw data. Grd applcatons such as astronomcal observatons, large-scale smulaton and sensor networks may requre workng n a brng-data-toprogram manner. There may be scenaros that owners of data, although ready for data sharng, may be reluctant to allow remote users to execute processng programs on stes for the sake of securty or lack of computatonal resources. For example, n the LIGO project, there are not enough computatonal resources avalable at observatores though data have to be generated from observatores. In order to utlze CPU cycles avalable from the OSG, LIGO data have to be streamed to OSG stes for processng. On the other hand, t s not always the case that transferrng large amount of data requres hgh bandwdth and local storage space. It s not necessarly ether that all data are transferred to local at one tme before the processng program s actually started. In ths case, data can be processed n the form of streams, and correspondng mplementaton requres supports of on-demand data transfers and just-n-tme data cleanups descrbed n the followng sectons On-demand data transfers It s not always the case that data should be transferred as fast as possble. For data streamng applcatons, data transfers should be on-demand nstead of spontaneous. Ths s to say, data transfers can be controlled to make requred data avalable accordng to actual data processng speeds. If data arrve too fast and cannot be processed n tme, accumulated data may requre large amount of storage space over tme; f data transfer speed s lower than processng, correspondng jobs may become dle and computatonal power s not fully utlzed. A data streamng envronment has to support multple applcatons, each requrng dfferent data types. All applcatons have to share lmt network bandwdth and storage space. For example, most OSG stes are CPU rch but storage lmt. Data transfers have to be scheduled n order to share bandwdth save local storage and maxmze CPU usage for data processng. The prevalng prncple here s enough s ok, not the faster transfers the better, nor the more data n storage the better. For data streamng applcatons, data transfers and processng take place smultaneously. The data streamng envronment ams to provde data for applcatons transparently and seamlessly as f jobs are processng data that were always avalable from local storage. Ths can be acheved va careful schedulng especally wth constrants of bandwdth and storage Just-n-tme data cleanups Besdes data transfer schedules, a cleanup procedure s ntegrated n the envronment to remove obsolete data and save storage space for subsequently comng data. Processed data should be removed from local storage as soon as possble. Just-n-tme cleanups are of great mportance n data streamng envronment snce the whole volume of data are usually extremely hgh. There maybe stuatons that multple applcatons are sharng a same set of data. In ths case, data can be only removed after all jobs are processed, otherwse the same set of data have to be transferred for multple tmes. 3. The Envronment Implementaton As shown n Fgure, mplementaton of the data streamng envronment s descrbed n ths secton, ncludng ndvdual processors for performance sensng, predcton, schedulng and executon of data transfers and cleanups.

3 Scheduler Predctor Data Transfer Job Manager Data Cleanup Sensor Sensor Sensor Storage Networks CPUs Globus GrdFTP Condor Fgure. The data streamng envronment Sensors. The grd s a dynamc envronment, where performance of resources, ncludng networks, processors and storage, are changng over tme. For example, the network s shared by many data transfers and the competton makes avalable bandwdth for each transfer vary from tme to tme, so real-tme performance nformaton s requred. Some sensors are developed to collect such nformaton, whch can be used as nputs of the performance predctor. Predctor. The predctor apples some predcton algorthms to forecast performance for a perod of tme n the future. Predcton results are nputs of the scheduler. Scheduler. The scheduler collects performance predcton results (e.g. data transferrng and processng speeds) and decdes when to transfer data, launch jobs and clean up processed data. Actual executons call external functons provded by Globus and Condor, as shown n Fgure. Our optmal goal s to make full use of computatonal resources, requrng guarantee of data provson, and mnmze bandwdth and storage consumptons. 3.. Real-tme measurements Performance nformaton s of mportance to predcton of both data transfers and processng. Ths can be acheved va real-tme performance measurements. Expermental results on GrdFTP scalablty are ncluded n ths secton as an example. We apply GrdFTP as the data transfer protocol for cross-doman data replcaton, parameters of whch can be adjusted to get optmal transferrng performance, ncludng parallelsm, TCP buffer sze and buffer sze. The parallelsm has the most drect mpact on data transfer speeds. Our experments show that the optmal rm number of data channels s between 8 and 0, as shown n Fgure 2. The curve stands for average tme of 20 experments for transferrng a data fle of 2 GB n seconds, usng dfferent parallelsms. It s obvous the data transfer speed ncreases dramatcally wth the GrdFTP parallelsm changng from to 8. When the parallelsm reaches over 0, the ncrement does not result n better performance further. It s a non-trval task to determne the proper amount of bandwdth to be allocated for each applcaton runnng n the Condor pool n terms of utlzaton and qualty of servce (QoS) satsfacton. As far as our assgnment algorthm concerns, t s transformed to set approprate GrdFTP parallelsms for applcatons. For convenence, the parallelsm s set to, 2, 4, 6 or 8 accordng to data processng speed and status of network. For example, f a certan processng program can consume data of 2 GB n 230 seconds, accordng to Fgure 2, the parallelsm can be set to 4 or 6 to guarantee data supply wth mnmum bandwdth. Average Transmsson Tme (seconds) Transmsson Parallelsm Fgure 2. Comparson of data transfer tmes usng dfferent GrdFTP parallelsms Snce the grd s a dynamc and resource sharng envronment, the same GrdFTP parameter may result n dfferent actual data transfer speed over tme, and data processng speed n the Condor pool may also vary over tme. A well set parallelsm may not always match data requrements over a long tme, whch makes t necessary to evaluate these parameters perodcally. In a tme nterval, f the GrdFTP can not supply enough data, ts parallelsm should be set to ts upper neghbors, e.g., from 6 to 8. But we must be cautous to set the parallelsm to a lower level unless the redundant transferrng speed s observed n several successve ntervals. If even the hghest parallelsm can not meet data processng requrements, the correspondng processor has to be nevtably dle and wat for more avalable data; f the data transferrng speed s hgh enough, some repertory polcy should be appled to avod data overflow.

4 3.2. Repertory polces A repertory strategy wth lower and upper lmts for each type of data s appled for the scheduler to decde the start and end of data transfers and ensure only reasonable local storage s requred. For most OSG stes, the total storage s lmt and must be shared by many types of data. On the other hand, t s meanngless to keep processed data. Repertory polces are defned to lower and upper lmts of storage space for each type of data. The lower lmt s used to guarantee that data processng can survve network collapse when no data can be streamed from sources to local storage for a certan perod of tme, whch mproves system robustness and ncreases CPU resource utlzaton. Ths s also llustrated n Secton 4 wth detaled expermental results. The upper lmt for each applcaton s used to guarantee that the overall amount of data n local storage does not exceed avalable storage space. Our strategy can be denoted as M(l, u ), where l and u stand for lower and upper lmts for data type (=,2,,n) and n s the total number of data types. Lower and upper lmts are manly used as thresholds to control start and end tmes of data transfers: when data amount scratches the lower lmt, more data should be transferred untl the amount reaches the upper lmt. Snce there are also data cleanups nvolved, data amount keeps changng and vares between lower and upper lmts Performance predcton Data transferrng and processng speeds must match each other so as to gan optmzed performance wth hgh throughput, so t s necessary to mplement realtme montorng and predcton. Onlne measurement and predcton of CPU usage s ncluded n ths secton. There are many avalable predcton methods, ncludng nonlnear tme-seres analyss, wavelet analyss, rough and fuzzy sets. Our mplement of fractal predcton for CPU usage can be calculated rapdly wth reasonable precson, whch can meet our requrement though not necessarly the best algorthm. The fractal dstrbuton can be descrbed as: C N =, D r where r s the sample tme, an ndependent varable; N s the usage percent of CPU, a varable correspondng to r; C s a constant to be calculated and D stands for the fractal dmenson. Defne a seres of ntal data N j (j=,,m) and the aggregate sum of th order can be calculated as: j Nk, = k=. S, j = j > S, k, k= The fractal dmenson D,j and the constant C,j can be calculated as: S, j+ rj D, j = ln ln S r, j D, j C, j S, j rj j+ =, where r j =j ( j=,,m-). Here the second order aggregated sum s appled snce for most cases ts predcton result s good enough and ts computng overhead s reasonable. In ths case the predcton s: D2, m S2, m * m N m+ = S D 2, m S, m 2, m ( m + ) Table. Fractal predcton of CPU usage (m=5) N j S,j S 2,j S 3,j S 4,j D,j D 2,j D 3,j D 4,j Predcted Measured Relatve % N 6 N 6 error One of predcton results s gven n Table. N j s an average CPU usage (%) durng a 0 seconds nterval. Whle D,j vares to some extent, D 2,j, D 3,j and D 4,j approach to a stable value respectvely. Ths means that fractal dmensons tend to be fxed and can be appled to make predcton. As shown n Table, the relatve error of the predcton N 6 s reasonably good. A predcaton of 00 samples s ncluded n Fgure 3. Whle there are some dfference between the predcton and measurements, predcton results does follow the trend of CPU usage measurements Schedulng of data streamng The amount of data n storage vares over tme and can be descrbed as the followng: Q () t = I () t d () t =,2,, n Q ( 0) = 0 =,2,, n, t, () t where Q () I and d () t stand for the varyng rate, assgned transferrng bandwdth and processng speed for data type (=,,n). () t s the dervatve of the Q

5 data amount n storage, whch reflects ntegrated effects of nput and cleanup of data. Data s streamed to local storage on-demand and then processed and removed. If I (t)= d (t), there s Q () t = 0, whch means that all data s processed and removed as soon as they are transferred, and theoretcally no extra storage s necessary at any tme. CPU Usage (%) predcton measurements Sample No. Fgure 3. Comparson of 00 predcton and measurements of CPU usage Data processng programs run constantly to process avalable data. If no data s locally avalable, they have to be dle and wat whch wastes computatonal resources. So d (t) can be descrbed as: 0 Q () t = 0 d () t = =,2,, n. > 0 Q () t > 0 Sometmes we can just take d (t) as a constant, and then we have: 0 Q () t = 0 d () t = =,2,, n. d Q () t > 0 As mentoned before, the total storage avalable s lmt and the constrant s: n = Q () t S, where S s the total avalable storage. As ntroduced n repertory polces, every type of data has ts lower and upper lmts on storage space, L and U : () t U =,2, n L Q, It s necessary to set values for lower and upper lmts for each data type snce these defne when to start and end transferrng data. In ths work these values are set manually, and eventually some heurstc methods can be appled to fnd more optmal values. I (t) represents data transferrng speeds and can be controlled to optmze performance n terms of throughput wth the followng constrant: n = I () t I() t, where I(t) s the total avalable local bandwdth at a gven tme t. In an actual envronment, I (t) can be adjustable usng carefully selected GrdFTP parallelsm p for each data type. Possble values of p are, 2, 4, 6, and 8 and can be selected as: ~, I I ( t,) ~ 2, I ( t,) < I I ( t,2) ~, p = 4, I ( t,2) < I I ( t,4) ~ 6, I ( t,4) < I I ( t,6) ~ 8, I > I ( t,6) where ~ I ( t ) represents the data transferrng requrement (n another word, data processng speed) for data type at a gven tme t. I (t,p) s the estmated data transferrng speed usng the GrdFTP parallelsm p for data type at a gven tme t. These can be measured and predcted usng methods proposed n Sectons 3. and 3.3. Wth above processes and constrants, the optmzaton goal s to mnmze the waste of computatonal resources, or n another word, to maxmze throughput of each data type: T f max d () t dt =,2,, n, 0 or the whole throughput for all types of data: n T f d () t dt T n max f = max d ()dt t, 0 = 0 = or, f each type of data processng has dfferent prvleges or weghts: n T f d () t dt T n max ω f = max ωd ()dt t 0 = 0 = where T f s the evaluaton tme span and ω s the weght of data type (=,,n). There could be many other schedulng goals, e.g. farness of data transferrng and processng. Many exstng polces n job schedulng can also be appled for schedulng of data streamng. These wll be addressed n our future work. 4. Expermental Results A prototype data streamng envronment s mplemented to evaluate algorthms proposed n Secton 3. A Condor pool s bult to manage over 0 workstatons n the laboratory, provdng computatonal and storage resources. Data to be processed are avalable n remote stes rather than machnes n the Condor pool, and Globus GrdFTP s adopted for data streamng. 4.. Experment desgn An LIGO data analyss program s used as a case study of data streamng applcatons. It reads n data

6 from both observatores and calculates statstcs for gravtatonal burst analyss. LIGO data for analyss are transformed from raw data to bnary data fles n reduced sze usng an nternatonal standard data format. Data processng programs are submtted to the Condor pool and allocated to avalable computatonal nodes accordng to ther requrements. Processng programs run constantly on host machnes, and requred data are transferred from data sources on remote stes to local storage and fed to correspondng processng programs. Our data streamng envronment wth several ndvdual daemons, ncludng sensors, the predctor, the scheduler and data transfer and cleanup agents, s ndependent from applcatons. It montors processng programs, collects nformaton about data transferrng and consumng to decde parameters dynamcally,.e. upper and lower lmts of repertory polces as thresholds of data transfers, GrdFTP parallelsms, and so on. Snce the grd s a dynamc envronment, schedulng s carred out over and over agan and makes new decsons when necessary, e.g. new applcatons jon or data processng speeds and network stuatons change Performance evaluaton - scalablty Three repertory polces are used for performance evaluaton of storage scalablty, whch are llustrated n Fgure 4. Contnuous streamng wthout cleanup. Data transfers are not under any control,.e. data are transferred constantly, and no cleanup procedure s nvolved, so data accumulates rapdly, as shown n Fgure 4. Contnuous streamng wth cleanup. The cleanup process s added. Whle the requred storage space scales better, data stll accumulates over tme. Ths s because data transfers are faster than processng. Although processed data can be removed mmedately, data watng to be processed stll accumulates and requred storage space ncreases. On-demand streamng wth cleanup. In ths scenaro, the schedulng algorthm proposed n Secton 3 s appled. Snce data transfers are controlled usng lower and upper lmts, data transfers would stop at upper lmts and data overflow would not happen. From expermental results gven n Fgure 4 we can nfer that our data streamng and schedulng mechansm should scale well wth applcatons requrng hgh volumes of data and long-term runs. Wth storage lower lmts, our repertory polces also result n tolerance to unexpected stuatons, e.g. network falures. Ths guarantees applcatons can survve network falures whch dsable data streamng from remote stes. Fgure 5 shows such an nstance. Storage Usage (MB) 2.5 x contnuous streamng wthout cleanup contnuous streamng wth cleanup on-demand streamng wth cleanup Tme (mnutes) Fgure 4. Performance evaluaton on storage scalablty usng 3 repertory strateges Durng a many hours data streamng perod, a network broken happens n the mddle for a short perod of tme. Whle the amount of local avalable data (represented usng data fle numbers) drops below the lower lmt temporarly, the applcaton s stll runnng and after the network recovers, data storage resumes to a normal level. Data Fle Number Tme (mnutes) Fgure 5. Storage usage tolerant to an unexpected network falure 4.3. Performance evaluaton - aglty We nvestgate aglty of our data streamng mplementaton by settng up an experment wth dynamcally changng data processng speeds. Ths could happen especally when multple processes are sharng CPU resources. Four typcal scenaros are desgned n the experment and a schedulng scheme of data streamng for 4 processng programs s llustrated n Fgure 6. data. Snce the data applcaton has a very hgh data processng speed, the GrdFTP parallelsm s set to the hghest 8 to meet the requrement. Even data has been streamed as fast as possble, t s obvous that data processng stll has to be dle and wat for avalable data occasonally.

7 data2. As shown n Fgure 6, the data processng speed of data2 applcaton decreases over tme, leadng to a dynamcally changng schedulng scheme for data transferrng. The GrdFTP parallelsm starts from 8 and fnally decreases to 2 snce processng speed becomes low after 4 mnutes. Processng Speed (MB/s) data data data 2 data 3 data Tme (mnutes) p=8 data2 p=8 p=6 p=6 p=2 5. Conclusons Ths work s focused on aglty aspects of data streamng for grd applcatons. On demand data transfers and just-n-tme data cleanups are proposed to mplement system aglty and meet applcaton requrements wth consderaton of scalablty. In order to utlze computatonal resources wth lmt storage and bandwdth, large amount of data have to be streamed for processng. A prototype of a data streamng envronment s mplemented, together wth Condor for CPU allocaton and Globus GrdFTP for data transfers. Some repertory polces are defned, performance measurement and predcton methods are proposed, and schedulng algorthms are mplemented. Expermental results show good scalablty and aglty of our envronment. Ongong work ncludes the consderaton of data sharng scenaros among multple data processng applcatons. Also some heurstc schedulng algorthm s under development for refned performance optmzaton. We are also nvestgatng the method to combne data streamng schedulng wth tradtonal job schedulng so that system aglty could be mproved further. data3 p=4 p=8 p=2 p=6 Acknowledgement data4 p= p= p= p= Tme (mnutes) Fgure 6. Performance evaluaton on aglty of data streamng to dynamcally changng data processng speeds data3. The data processng speed of data3 applcaton changes dramatcally over tme. The data streamng envronment has to adjust the GrdFTP parallelsm p on-the-fly between 2 and 8. data4. In contrast wth data applcaton, data4 s slow so p= can always satsfy the processng requrement, even wth some ntervals when no data transfers happen. In all of above scenaros except data, data transfers start and stop automatcally, meetng applcaton requrements as well as avodng data overflows. In ths work, we provde an ntegrated grd resource schedulng scheme for data streamng applcatons wth combnaton of consderatons of relevant resources, e.g. CPUs, storage and bandwdth. Ths mproves the system aglty to serve ndvdual data processes n a customzed and personalzed way, though applcaton requrements are dynamcally changng. Ths work s supported by Mnstry of Educaton of Chna under the 2/5 project Natonal Unversty Instrument and Resource Sharng Systems, and Mnstry of Scence and Technology of Chna under the natonal 863 hgh-tech R&D program (grants No. 2006AA0Z237, No. 2007AA0Z79 and No. 2008AA0Z8). Junwe Cao would lke to express hs grattude to Prof. Erk Katsavounds of LIGO (Laser Interferometer Gravtatonal-wave Observatory) Laboratory at Massachusetts Insttute of Technology for ther long-term collaboraton supports. References []. B. Agarwalla, N. Ahmed, D. Hlley, and U. Ramachandran, Streamlne: a Schedulng Heurstc for Streamng Applcatons on the Grd, n Proc. SPIE Multmeda Computng and Networkng, Vol. 607, [2]. B. Agarwalla, N. Ahmed, D. Hlley, and U. Ramachandran, Streamlne: Schedulng Streamng Applcatons n a Wde Area Envronment, Multmeda Systems, Vol. 3, No., pp , [3]. B. Allcock, J. Bester, J. Bresnahan, A. L. Chervenak, I. Foster, C. Kesselman, S. Meder, V. Nefedova, D. Quesnal, and S. Tuecke, Data Management and

8 Transfer n Hgh Performance Computatonal Grd Envronments, Parallel Computng, Vol. 28, No. 5, pp , [4]. A. Arasu, B. Babcock, S. Babu, M. Datar, K. Ito, I. Nshzawa, J. Rosensten, and J. Wdom, STREAM: The Stanford Stream Data Manager, IEEE Data Eng. Bull., Vol. 26, No., pp.9-26, [5]. V. Bhat, S. Klasky, S. Atchley, M. Beck, D. McCune, and M. Parashar, Hgh Performance Threaded Data Streamng for Large Scale Smulatons, n Proc. 5 th IEEE/ACM Int. Workshop on Grd Computng, pp , [6]. D. A. Brown, P. R. Brady, A. Detz, J. Cao, B. Johnson, and J. McNabb, A Case Study on the Use of Workflow Technologes for Scentfc Analyss: Gravtatonal Wave Data Analyss, n I. J. Taylor, D. Gannon, E. Deelman, and M. S. Shelds (Eds.), Workflows for escence: Scentfc Workflows for Grds, Sprnger Verlag, [7]. D. Carney, U. Cetnterne, M. Chernack, et. al., Montorng Streams: A New Class of Data Management Applcatons, n Proc. Int. Conf. on Very Large Data Bases, pp , [8]. J. Chen, D. DeWtt, F. Tan, and Y. Wang, NagaraCQ: A Scalable Contnuous Query System for Internet Databases, n Proc. ACM SIGMOD Int. Conf. on Management of Data, pp , [9]. A. Cockburn, Agle Software Development, Addson- Wesley, 200. [0]. C. Cranoe, T. Johnson, V. Shkapenyuk, and O. Spatscheck, Ggascope: a Stream Database for Network Applcatons, n Proc. ACM SIGMOD Int. Conf. on Management of Data, pp , []. S. Davs, and B. J. Pne II, Mass Customzaton: The New Fronter n Busness Competton, Harvard Busness School Press, 999. [2]. E. Deelman, J. Blythe, Y. Gl, C. Kesselman, G. Mehta, K. Vah, et. al., Mappng Abstract Complex Workflows onto Grd Envronments, J. Grd Computng, Vol., No., pp , [3]. E. Deelman, C. Kesselman, G. Mehta, L. Meshkat, L. Pearlman, et. al., GrPhyN and LIGO, Buldng a Vrtual Data Grd for Gravtatonal Wave Scentsts, n Proc. th IEEE Int. Symp. on Hgh Performance Dstrbuted Computng, pp , [4]. I. Foster and C. Kesselman, Globus: A Metacomputng Infrastructure Toolkt, Int. J. Supercomputer Applcatons, Vol., No. 2, pp. 5-28, 997. [5]. G. Fox, H. Gadgl, S. Pallckara, M. Perce, R. L. Grossman, et. al., Hgh Performance Data Streamng n Servce Archtecture, Techncal Report, Indana Unversty and Unversty of Illnos at Chcago, July 2004 [6]. P. T. Kdd, Agle Manufacturng: Forgng New Fronters, Addson-Wesley, 994. [7]. S. Klasky, S. Ether, Z. Ln, K. Martns, D. McCune and R. Samtaney, Grd-Based Parallel Data Streamng Implemented for the Gyroknetc Torodal Code, n Proc. ACM/IEEE Supercomputng Conf., [8]. R. Kuntschke, T. Scholl, S. Huber, A. Kemper, A. Reser, et. al, Grd-Based Data Stream Processng n e-scence, n Proc. 2 nd IEEE Int. Conf. on e-scence and Grd Computng, [9]. M. Ltzkow, M. Lvny, and Matt Mutka, Condor A Hunter of Idle Workstatons, n Proc. of 8 th Int. Conf. on Dstrbuted Computng Systems, pp. 04-, 988. [20]. R. Pordes for the Open Scence Grd Consortum, The Open Scence Grd, n Proc. Computng n Hgh Energy and Nuclear Physcs Conf., Interlaken, Swtzerland, [2]. A. Ramakrshnan, G. Sngh, H. Zhao, E. Deelman, R. Sakellarou, K. Vah, K. Blackburn, D. Meyers, and M. Samd, Schedulng Data-ntensve Workflow onto Storage-Constraned Dstrbuted Resources, n Proc. 7 th IEEE Int. Symp. on Cluster Computng and the Grd, Ro de Janero, Brazl, pp , [22]. Y. Zhu and D. Shasha, Statstream: Statstcal Montorng of Thousands of Data Streams n Real Tme, Techncal Report TR , CS Dept, New York Unversty, 2002.

Scheduling Remote Access to Scientific Instruments in Cyberinfrastructure for Education and Research

Scheduling Remote Access to Scientific Instruments in Cyberinfrastructure for Education and Research Schedulng Remote Access to Scentfc Instruments n Cybernfrastructure for Educaton and Research Je Yn 1, Junwe Cao 2,3,*, Yuexuan Wang 4, Lanchen Lu 1,3 and Cheng Wu 1,3 1 Natonal CIMS Engneerng and Research

More information

A Binarization Algorithm specialized on Document Images and Photos

A Binarization Algorithm specialized on Document Images and Photos A Bnarzaton Algorthm specalzed on Document mages and Photos Ergna Kavalleratou Dept. of nformaton and Communcaton Systems Engneerng Unversty of the Aegean kavalleratou@aegean.gr Abstract n ths paper, a

More information

Virtual Memory. Background. No. 10. Virtual Memory: concept. Logical Memory Space (review) Demand Paging(1) Virtual Memory

Virtual Memory. Background. No. 10. Virtual Memory: concept. Logical Memory Space (review) Demand Paging(1) Virtual Memory Background EECS. Operatng System Fundamentals No. Vrtual Memory Prof. Hu Jang Department of Electrcal Engneerng and Computer Scence, York Unversty Memory-management methods normally requres the entre process

More information

Simulation Based Analysis of FAST TCP using OMNET++

Simulation Based Analysis of FAST TCP using OMNET++ Smulaton Based Analyss of FAST TCP usng OMNET++ Umar ul Hassan 04030038@lums.edu.pk Md Term Report CS678 Topcs n Internet Research Sprng, 2006 Introducton Internet traffc s doublng roughly every 3 months

More information

Problem Definitions and Evaluation Criteria for Computational Expensive Optimization

Problem Definitions and Evaluation Criteria for Computational Expensive Optimization Problem efntons and Evaluaton Crtera for Computatonal Expensve Optmzaton B. Lu 1, Q. Chen and Q. Zhang 3, J. J. Lang 4, P. N. Suganthan, B. Y. Qu 6 1 epartment of Computng, Glyndwr Unversty, UK Faclty

More information

Distributed Resource Scheduling in Grid Computing Using Fuzzy Approach

Distributed Resource Scheduling in Grid Computing Using Fuzzy Approach Dstrbuted Resource Schedulng n Grd Computng Usng Fuzzy Approach Shahram Amn, Mohammad Ahmad Computer Engneerng Department Islamc Azad Unversty branch Mahallat, Iran Islamc Azad Unversty branch khomen,

More information

Video Proxy System for a Large-scale VOD System (DINA)

Video Proxy System for a Large-scale VOD System (DINA) Vdeo Proxy System for a Large-scale VOD System (DINA) KWUN-CHUNG CHAN #, KWOK-WAI CHEUNG *# #Department of Informaton Engneerng *Centre of Innovaton and Technology The Chnese Unversty of Hong Kong SHATIN,

More information

A mathematical programming approach to the analysis, design and scheduling of offshore oilfields

A mathematical programming approach to the analysis, design and scheduling of offshore oilfields 17 th European Symposum on Computer Aded Process Engneerng ESCAPE17 V. Plesu and P.S. Agach (Edtors) 2007 Elsever B.V. All rghts reserved. 1 A mathematcal programmng approach to the analyss, desgn and

More information

Wishing you all a Total Quality New Year!

Wishing you all a Total Quality New Year! Total Qualty Management and Sx Sgma Post Graduate Program 214-15 Sesson 4 Vnay Kumar Kalakband Assstant Professor Operatons & Systems Area 1 Wshng you all a Total Qualty New Year! Hope you acheve Sx sgma

More information

Virtual Machine Migration based on Trust Measurement of Computer Node

Virtual Machine Migration based on Trust Measurement of Computer Node Appled Mechancs and Materals Onlne: 2014-04-04 ISSN: 1662-7482, Vols. 536-537, pp 678-682 do:10.4028/www.scentfc.net/amm.536-537.678 2014 Trans Tech Publcatons, Swtzerland Vrtual Machne Mgraton based on

More information

Parallelism for Nested Loops with Non-uniform and Flow Dependences

Parallelism for Nested Loops with Non-uniform and Flow Dependences Parallelsm for Nested Loops wth Non-unform and Flow Dependences Sam-Jn Jeong Dept. of Informaton & Communcaton Engneerng, Cheonan Unversty, 5, Anseo-dong, Cheonan, Chungnam, 330-80, Korea. seong@cheonan.ac.kr

More information

Compiler Design. Spring Register Allocation. Sample Exercises and Solutions. Prof. Pedro C. Diniz

Compiler Design. Spring Register Allocation. Sample Exercises and Solutions. Prof. Pedro C. Diniz Compler Desgn Sprng 2014 Regster Allocaton Sample Exercses and Solutons Prof. Pedro C. Dnz USC / Informaton Scences Insttute 4676 Admralty Way, Sute 1001 Marna del Rey, Calforna 90292 pedro@s.edu Regster

More information

An Optimal Algorithm for Prufer Codes *

An Optimal Algorithm for Prufer Codes * J. Software Engneerng & Applcatons, 2009, 2: 111-115 do:10.4236/jsea.2009.22016 Publshed Onlne July 2009 (www.scrp.org/journal/jsea) An Optmal Algorthm for Prufer Codes * Xaodong Wang 1, 2, Le Wang 3,

More information

Improved Resource Allocation Algorithms for Practical Image Encoding in a Ubiquitous Computing Environment

Improved Resource Allocation Algorithms for Practical Image Encoding in a Ubiquitous Computing Environment JOURNAL OF COMPUTERS, VOL. 4, NO. 9, SEPTEMBER 2009 873 Improved Resource Allocaton Algorthms for Practcal Image Encodng n a Ubqutous Computng Envronment Manxong Dong, Long Zheng, Kaoru Ota, Song Guo School

More information

A New Token Allocation Algorithm for TCP Traffic in Diffserv Network

A New Token Allocation Algorithm for TCP Traffic in Diffserv Network A New Token Allocaton Algorthm for TCP Traffc n Dffserv Network A New Token Allocaton Algorthm for TCP Traffc n Dffserv Network S. Sudha and N. Ammasagounden Natonal Insttute of Technology, Truchrappall,

More information

Efficient Distributed File System (EDFS)

Efficient Distributed File System (EDFS) Effcent Dstrbuted Fle System (EDFS) (Sem-Centralzed) Debessay(Debsh) Fesehaye, Rahul Malk & Klara Naherstedt Unversty of Illnos-Urbana Champagn Contents Problem Statement, Related Work, EDFS Desgn Rate

More information

Cluster Analysis of Electrical Behavior

Cluster Analysis of Electrical Behavior Journal of Computer and Communcatons, 205, 3, 88-93 Publshed Onlne May 205 n ScRes. http://www.scrp.org/ournal/cc http://dx.do.org/0.4236/cc.205.350 Cluster Analyss of Electrcal Behavor Ln Lu Ln Lu, School

More information

Resource and Virtual Function Status Monitoring in Network Function Virtualization Environment

Resource and Virtual Function Status Monitoring in Network Function Virtualization Environment Journal of Physcs: Conference Seres PAPER OPEN ACCESS Resource and Vrtual Functon Status Montorng n Network Functon Vrtualzaton Envronment To cte ths artcle: MS Ha et al 2018 J. Phys.: Conf. Ser. 1087

More information

Real-time Fault-tolerant Scheduling Algorithm for Distributed Computing Systems

Real-time Fault-tolerant Scheduling Algorithm for Distributed Computing Systems Real-tme Fault-tolerant Schedulng Algorthm for Dstrbuted Computng Systems Yun Lng, Y Ouyang College of Computer Scence and Informaton Engneerng Zheang Gongshang Unversty Postal code: 310018 P.R.CHINA {ylng,

More information

Learning the Kernel Parameters in Kernel Minimum Distance Classifier

Learning the Kernel Parameters in Kernel Minimum Distance Classifier Learnng the Kernel Parameters n Kernel Mnmum Dstance Classfer Daoqang Zhang 1,, Songcan Chen and Zh-Hua Zhou 1* 1 Natonal Laboratory for Novel Software Technology Nanjng Unversty, Nanjng 193, Chna Department

More information

Cost-efficient deployment of distributed software services

Cost-efficient deployment of distributed software services 1/30 Cost-effcent deployment of dstrbuted software servces csorba@tem.ntnu.no 2/30 Short ntroducton & contents Cost-effcent deployment of dstrbuted software servces Cost functons Bo-nspred decentralzed

More information

Determining the Optimal Bandwidth Based on Multi-criterion Fusion

Determining the Optimal Bandwidth Based on Multi-criterion Fusion Proceedngs of 01 4th Internatonal Conference on Machne Learnng and Computng IPCSIT vol. 5 (01) (01) IACSIT Press, Sngapore Determnng the Optmal Bandwdth Based on Mult-crteron Fuson Ha-L Lang 1+, Xan-Mn

More information

Comparison of Heuristics for Scheduling Independent Tasks on Heterogeneous Distributed Environments

Comparison of Heuristics for Scheduling Independent Tasks on Heterogeneous Distributed Environments Comparson of Heurstcs for Schedulng Independent Tasks on Heterogeneous Dstrbuted Envronments Hesam Izakan¹, Ath Abraham², Senor Member, IEEE, Václav Snášel³ ¹ Islamc Azad Unversty, Ramsar Branch, Ramsar,

More information

A MOVING MESH APPROACH FOR SIMULATION BUDGET ALLOCATION ON CONTINUOUS DOMAINS

A MOVING MESH APPROACH FOR SIMULATION BUDGET ALLOCATION ON CONTINUOUS DOMAINS Proceedngs of the Wnter Smulaton Conference M E Kuhl, N M Steger, F B Armstrong, and J A Jones, eds A MOVING MESH APPROACH FOR SIMULATION BUDGET ALLOCATION ON CONTINUOUS DOMAINS Mark W Brantley Chun-Hung

More information

Two-Stage Data Distribution for Distributed Surveillance Video Processing with Hybrid Storage Architecture

Two-Stage Data Distribution for Distributed Surveillance Video Processing with Hybrid Storage Architecture Two-Stage Data Dstrbuton for Dstrbuted Survellance Vdeo Processng wth Hybrd Storage Archtecture Yangyang Gao, Hatao Zhang, Bngchang Tang, Yanpe Zhu, Huadong Ma Bejng Key Lab of Intellgent Telecomm. Software

More information

An Entropy-Based Approach to Integrated Information Needs Assessment

An Entropy-Based Approach to Integrated Information Needs Assessment Dstrbuton Statement A: Approved for publc release; dstrbuton s unlmted. An Entropy-Based Approach to ntegrated nformaton Needs Assessment June 8, 2004 Wllam J. Farrell Lockheed Martn Advanced Technology

More information

X- Chart Using ANOM Approach

X- Chart Using ANOM Approach ISSN 1684-8403 Journal of Statstcs Volume 17, 010, pp. 3-3 Abstract X- Chart Usng ANOM Approach Gullapall Chakravarth 1 and Chaluvad Venkateswara Rao Control lmts for ndvdual measurements (X) chart are

More information

AADL : about scheduling analysis

AADL : about scheduling analysis AADL : about schedulng analyss Schedulng analyss, what s t? Embedded real-tme crtcal systems have temporal constrants to meet (e.g. deadlne). Many systems are bult wth operatng systems provdng multtaskng

More information

NUMERICAL SOLVING OPTIMAL CONTROL PROBLEMS BY THE METHOD OF VARIATIONS

NUMERICAL SOLVING OPTIMAL CONTROL PROBLEMS BY THE METHOD OF VARIATIONS ARPN Journal of Engneerng and Appled Scences 006-017 Asan Research Publshng Network (ARPN). All rghts reserved. NUMERICAL SOLVING OPTIMAL CONTROL PROBLEMS BY THE METHOD OF VARIATIONS Igor Grgoryev, Svetlana

More information

The Codesign Challenge

The Codesign Challenge ECE 4530 Codesgn Challenge Fall 2007 Hardware/Software Codesgn The Codesgn Challenge Objectves In the codesgn challenge, your task s to accelerate a gven software reference mplementaton as fast as possble.

More information

Fast Retransmission of Real-Time Traffic in HIPERLAN/2 Systems

Fast Retransmission of Real-Time Traffic in HIPERLAN/2 Systems Fast Retransmsson of Real-Tme Traffc n HIPERLAN/ Systems José A Afonso and Joaqum E Neves Department of Industral Electroncs Unversty of Mnho, Campus de Azurém 4800-058 Gumarães, Portugal {joseafonso,

More information

Adaptive Energy and Location Aware Routing in Wireless Sensor Network

Adaptive Energy and Location Aware Routing in Wireless Sensor Network Adaptve Energy and Locaton Aware Routng n Wreless Sensor Network Hong Fu 1,1, Xaomng Wang 1, Yngshu L 1 Department of Computer Scence, Shaanx Normal Unversty, X an, Chna, 71006 fuhong433@gmal.com {wangxmsnnu@hotmal.cn}

More information

Concurrent Apriori Data Mining Algorithms

Concurrent Apriori Data Mining Algorithms Concurrent Apror Data Mnng Algorthms Vassl Halatchev Department of Electrcal Engneerng and Computer Scence York Unversty, Toronto October 8, 2015 Outlne Why t s mportant Introducton to Assocaton Rule Mnng

More information

Distributed Middlebox Placement Based on Potential Game

Distributed Middlebox Placement Based on Potential Game Int. J. Communcatons, Network and System Scences, 2017, 10, 264-273 http://www.scrp.org/ournal/cns ISSN Onlne: 1913-3723 ISSN Prnt: 1913-3715 Dstrbuted Mddlebox Placement Based on Potental Game Yongwen

More information

S1 Note. Basis functions.

S1 Note. Basis functions. S1 Note. Bass functons. Contents Types of bass functons...1 The Fourer bass...2 B-splne bass...3 Power and type I error rates wth dfferent numbers of bass functons...4 Table S1. Smulaton results of type

More information

Learning-Based Top-N Selection Query Evaluation over Relational Databases

Learning-Based Top-N Selection Query Evaluation over Relational Databases Learnng-Based Top-N Selecton Query Evaluaton over Relatonal Databases Lang Zhu *, Wey Meng ** * School of Mathematcs and Computer Scence, Hebe Unversty, Baodng, Hebe 071002, Chna, zhu@mal.hbu.edu.cn **

More information

Helsinki University Of Technology, Systems Analysis Laboratory Mat Independent research projects in applied mathematics (3 cr)

Helsinki University Of Technology, Systems Analysis Laboratory Mat Independent research projects in applied mathematics (3 cr) Helsnk Unversty Of Technology, Systems Analyss Laboratory Mat-2.08 Independent research projects n appled mathematcs (3 cr) "! #$&% Antt Laukkanen 506 R ajlaukka@cc.hut.f 2 Introducton...3 2 Multattrbute

More information

Term Weighting Classification System Using the Chi-square Statistic for the Classification Subtask at NTCIR-6 Patent Retrieval Task

Term Weighting Classification System Using the Chi-square Statistic for the Classification Subtask at NTCIR-6 Patent Retrieval Task Proceedngs of NTCIR-6 Workshop Meetng, May 15-18, 2007, Tokyo, Japan Term Weghtng Classfcaton System Usng the Ch-square Statstc for the Classfcaton Subtask at NTCIR-6 Patent Retreval Task Kotaro Hashmoto

More information

An Iterative Solution Approach to Process Plant Layout using Mixed Integer Optimisation

An Iterative Solution Approach to Process Plant Layout using Mixed Integer Optimisation 17 th European Symposum on Computer Aded Process Engneerng ESCAPE17 V. Plesu and P.S. Agach (Edtors) 2007 Elsever B.V. All rghts reserved. 1 An Iteratve Soluton Approach to Process Plant Layout usng Mxed

More information

Load-Balanced Anycast Routing

Load-Balanced Anycast Routing Load-Balanced Anycast Routng Chng-Yu Ln, Jung-Hua Lo, and Sy-Yen Kuo Department of Electrcal Engneerng atonal Tawan Unversty, Tape, Tawan sykuo@cc.ee.ntu.edu.tw Abstract For fault-tolerance and load-balance

More information

User Authentication Based On Behavioral Mouse Dynamics Biometrics

User Authentication Based On Behavioral Mouse Dynamics Biometrics User Authentcaton Based On Behavoral Mouse Dynamcs Bometrcs Chee-Hyung Yoon Danel Donghyun Km Department of Computer Scence Department of Computer Scence Stanford Unversty Stanford Unversty Stanford, CA

More information

A Fast Visual Tracking Algorithm Based on Circle Pixels Matching

A Fast Visual Tracking Algorithm Based on Circle Pixels Matching A Fast Vsual Trackng Algorthm Based on Crcle Pxels Matchng Zhqang Hou hou_zhq@sohu.com Chongzhao Han czhan@mal.xjtu.edu.cn Ln Zheng Abstract: A fast vsual trackng algorthm based on crcle pxels matchng

More information

Related-Mode Attacks on CTR Encryption Mode

Related-Mode Attacks on CTR Encryption Mode Internatonal Journal of Network Securty, Vol.4, No.3, PP.282 287, May 2007 282 Related-Mode Attacks on CTR Encrypton Mode Dayn Wang, Dongda Ln, and Wenlng Wu (Correspondng author: Dayn Wang) Key Laboratory

More information

Support Vector Machines

Support Vector Machines /9/207 MIST.6060 Busness Intellgence and Data Mnng What are Support Vector Machnes? Support Vector Machnes Support Vector Machnes (SVMs) are supervsed learnng technques that analyze data and recognze patterns.

More information

Module Management Tool in Software Development Organizations

Module Management Tool in Software Development Organizations Journal of Computer Scence (5): 8-, 7 ISSN 59-66 7 Scence Publcatons Management Tool n Software Development Organzatons Ahmad A. Al-Rababah and Mohammad A. Al-Rababah Faculty of IT, Al-Ahlyyah Amman Unversty,

More information

Reliability and Performance Models for Grid Computing

Reliability and Performance Models for Grid Computing Relablty and Performance Models for Grd Computng Yuan-Shun Da,2, Jack Dongarra,3,4 Department of Electrcal Engneerng and Computer Scence, Unversty of Tennessee, Knoxvlle 2 Department of Industral and Informaton

More information

Dynamic Bandwidth Provisioning with Fairness and Revenue Considerations for Broadband Wireless Communication

Dynamic Bandwidth Provisioning with Fairness and Revenue Considerations for Broadband Wireless Communication Ths full text paper was peer revewed at the drecton of IEEE Communcatons Socety subject matter experts for publcaton n the ICC 008 proceedngs. Dynamc Bandwdth Provsonng wth Farness and Revenue Consderatons

More information

RAP. Speed/RAP/CODA. Real-time Systems. Modeling the sensor networks. Real-time Systems. Modeling the sensor networks. Real-time systems:

RAP. Speed/RAP/CODA. Real-time Systems. Modeling the sensor networks. Real-time Systems. Modeling the sensor networks. Real-time systems: Speed/RAP/CODA Presented by Octav Chpara Real-tme Systems Many wreless sensor network applcatons requre real-tme support Survellance and trackng Border patrol Fre fghtng Real-tme systems: Hard real-tme:

More information

USING GRAPHING SKILLS

USING GRAPHING SKILLS Name: BOLOGY: Date: _ Class: USNG GRAPHNG SKLLS NTRODUCTON: Recorded data can be plotted on a graph. A graph s a pctoral representaton of nformaton recorded n a data table. t s used to show a relatonshp

More information

Motivation. EE 457 Unit 4. Throughput vs. Latency. Performance Depends on View Point?! Computer System Performance. An individual user wants to:

Motivation. EE 457 Unit 4. Throughput vs. Latency. Performance Depends on View Point?! Computer System Performance. An individual user wants to: 4.1 4.2 Motvaton EE 457 Unt 4 Computer System Performance An ndvdual user wants to: Mnmze sngle program executon tme A datacenter owner wants to: Maxmze number of Mnmze ( ) http://e-tellgentnternetmarketng.com/webste/frustrated-computer-user-2/

More information

A QoS-aware Scheduling Scheme for Software-Defined Storage Oriented iscsi Target

A QoS-aware Scheduling Scheme for Software-Defined Storage Oriented iscsi Target A QoS-aware Schedulng Scheme for Software-Defned Storage Orented SCSI Target Xanghu Meng 1,2, Xuewen Zeng 1, Xao Chen 1, Xaozhou Ye 1,* 1 Natonal Network New Meda Engneerng Research Center, Insttute of

More information

Application of Improved Fish Swarm Algorithm in Cloud Computing Resource Scheduling

Application of Improved Fish Swarm Algorithm in Cloud Computing Resource Scheduling , pp.40-45 http://dx.do.org/10.14257/astl.2017.143.08 Applcaton of Improved Fsh Swarm Algorthm n Cloud Computng Resource Schedulng Yu Lu, Fangtao Lu School of Informaton Engneerng, Chongqng Vocatonal Insttute

More information

FAHP and Modified GRA Based Network Selection in Heterogeneous Wireless Networks

FAHP and Modified GRA Based Network Selection in Heterogeneous Wireless Networks 2017 2nd Internatonal Semnar on Appled Physcs, Optoelectroncs and Photoncs (APOP 2017) ISBN: 978-1-60595-522-3 FAHP and Modfed GRA Based Network Selecton n Heterogeneous Wreless Networks Xaohan DU, Zhqng

More information

Research of Dynamic Access to Cloud Database Based on Improved Pheromone Algorithm

Research of Dynamic Access to Cloud Database Based on Improved Pheromone Algorithm , pp.197-202 http://dx.do.org/10.14257/dta.2016.9.5.20 Research of Dynamc Access to Cloud Database Based on Improved Pheromone Algorthm Yongqang L 1 and Jn Pan 2 1 (Software Technology Vocatonal College,

More information

Meta-heuristics for Multidimensional Knapsack Problems

Meta-heuristics for Multidimensional Knapsack Problems 2012 4th Internatonal Conference on Computer Research and Development IPCSIT vol.39 (2012) (2012) IACSIT Press, Sngapore Meta-heurstcs for Multdmensonal Knapsack Problems Zhbao Man + Computer Scence Department,

More information

ARTICLE IN PRESS. Signal Processing: Image Communication

ARTICLE IN PRESS. Signal Processing: Image Communication Sgnal Processng: Image Communcaton 23 (2008) 754 768 Contents lsts avalable at ScenceDrect Sgnal Processng: Image Communcaton journal homepage: www.elsever.com/locate/mage Dstrbuted meda rate allocaton

More information

Solving two-person zero-sum game by Matlab

Solving two-person zero-sum game by Matlab Appled Mechancs and Materals Onlne: 2011-02-02 ISSN: 1662-7482, Vols. 50-51, pp 262-265 do:10.4028/www.scentfc.net/amm.50-51.262 2011 Trans Tech Publcatons, Swtzerland Solvng two-person zero-sum game by

More information

Application of VCG in Replica Placement Strategy of Cloud Storage

Application of VCG in Replica Placement Strategy of Cloud Storage Internatonal Journal of Grd and Dstrbuted Computng, pp.27-40 http://dx.do.org/10.14257/jgdc.2016.9.4.03 Applcaton of VCG n Replca Placement Strategy of Cloud Storage Wang Hongxa Computer Department, Bejng

More information

Hybrid Job Scheduling Mechanism Using a Backfill-based Multi-queue Strategy in Distributed Grid Computing

Hybrid Job Scheduling Mechanism Using a Backfill-based Multi-queue Strategy in Distributed Grid Computing IJCSNS Internatonal Journal of Computer Scence and Network Securty, VOL.12 No.9, September 2012 39 Hybrd Job Schedulng Mechansm Usng a Backfll-based Mult-queue Strategy n Dstrbuted Grd Computng Ken Park

More information

Analysis of Collaborative Distributed Admission Control in x Networks

Analysis of Collaborative Distributed Admission Control in x Networks 1 Analyss of Collaboratve Dstrbuted Admsson Control n 82.11x Networks Thnh Nguyen, Member, IEEE, Ken Nguyen, Member, IEEE, Lnha He, Member, IEEE, Abstract Wth the recent surge of wreless home networks,

More information

Constructing Minimum Connected Dominating Set: Algorithmic approach

Constructing Minimum Connected Dominating Set: Algorithmic approach Constructng Mnmum Connected Domnatng Set: Algorthmc approach G.N. Puroht and Usha Sharma Centre for Mathematcal Scences, Banasthal Unversty, Rajasthan 304022 usha.sharma94@yahoo.com Abstract: Connected

More information

TECHNIQUE OF FORMATION HOMOGENEOUS SAMPLE SAME OBJECTS. Muradaliyev A.Z.

TECHNIQUE OF FORMATION HOMOGENEOUS SAMPLE SAME OBJECTS. Muradaliyev A.Z. TECHNIQUE OF FORMATION HOMOGENEOUS SAMPLE SAME OBJECTS Muradalyev AZ Azerbajan Scentfc-Research and Desgn-Prospectng Insttute of Energetc AZ1012, Ave HZardab-94 E-mal:aydn_murad@yahoocom Importance of

More information

A Genetic Algorithm Based Dynamic Load Balancing Scheme for Heterogeneous Distributed Systems

A Genetic Algorithm Based Dynamic Load Balancing Scheme for Heterogeneous Distributed Systems Proceedngs of the Internatonal Conference on Parallel and Dstrbuted Processng Technques and Applcatons, PDPTA 2008, Las Vegas, Nevada, USA, July 14-17, 2008, 2 Volumes. CSREA Press 2008, ISBN 1-60132-084-1

More information

Run-Time Operator State Spilling for Memory Intensive Long-Running Queries

Run-Time Operator State Spilling for Memory Intensive Long-Running Queries Run-Tme Operator State Spllng for Memory Intensve Long-Runnng Queres Bn Lu, Yal Zhu, and lke A. Rundenstener epartment of Computer Scence, Worcester Polytechnc Insttute Worcester, Massachusetts, USA {bnlu,

More information

CS 534: Computer Vision Model Fitting

CS 534: Computer Vision Model Fitting CS 534: Computer Vson Model Fttng Sprng 004 Ahmed Elgammal Dept of Computer Scence CS 534 Model Fttng - 1 Outlnes Model fttng s mportant Least-squares fttng Maxmum lkelhood estmaton MAP estmaton Robust

More information

Channel 0. Channel 1 Channel 2. Channel 3 Channel 4. Channel 5 Channel 6 Channel 7

Channel 0. Channel 1 Channel 2. Channel 3 Channel 4. Channel 5 Channel 6 Channel 7 Optmzed Regonal Cachng for On-Demand Data Delvery Derek L. Eager Mchael C. Ferrs Mary K. Vernon Unversty of Saskatchewan Unversty of Wsconsn Madson Saskatoon, SK Canada S7N 5A9 Madson, WI 5376 eager@cs.usask.ca

More information

Classifier Selection Based on Data Complexity Measures *

Classifier Selection Based on Data Complexity Measures * Classfer Selecton Based on Data Complexty Measures * Edth Hernández-Reyes, J.A. Carrasco-Ochoa, and J.Fco. Martínez-Trndad Natonal Insttute for Astrophyscs, Optcs and Electroncs, Lus Enrque Erro No.1 Sta.

More information

An Approach to Optimized Resource Scheduling Algorithm for Open-source Cloud Systems

An Approach to Optimized Resource Scheduling Algorithm for Open-source Cloud Systems The Ffth Annual ChnaGrd Conference An Approach to Optmzed Resource Schedulng Algorthm for Open-source Cloud Systems Ha Zhong 1, 2, Kun Tao 1, Xueje Zhang 1, 2 1 School of Informaton Scence and Engneerng,

More information

Internet Traffic Managers

Internet Traffic Managers Internet Traffc Managers Ibrahm Matta matta@cs.bu.edu www.cs.bu.edu/faculty/matta Computer Scence Department Boston Unversty Boston, MA 225 Jont work wth members of the WING group: Azer Bestavros, John

More information

The Research of Support Vector Machine in Agricultural Data Classification

The Research of Support Vector Machine in Agricultural Data Classification The Research of Support Vector Machne n Agrcultural Data Classfcaton Le Sh, Qguo Duan, Xnmng Ma, Me Weng College of Informaton and Management Scence, HeNan Agrcultural Unversty, Zhengzhou 45000 Chna Zhengzhou

More information

Assignment # 2. Farrukh Jabeen Algorithms 510 Assignment #2 Due Date: June 15, 2009.

Assignment # 2. Farrukh Jabeen Algorithms 510 Assignment #2 Due Date: June 15, 2009. Farrukh Jabeen Algorthms 51 Assgnment #2 Due Date: June 15, 29. Assgnment # 2 Chapter 3 Dscrete Fourer Transforms Implement the FFT for the DFT. Descrbed n sectons 3.1 and 3.2. Delverables: 1. Concse descrpton

More information

Kent State University CS 4/ Design and Analysis of Algorithms. Dept. of Math & Computer Science LECT-16. Dynamic Programming

Kent State University CS 4/ Design and Analysis of Algorithms. Dept. of Math & Computer Science LECT-16. Dynamic Programming CS 4/560 Desgn and Analyss of Algorthms Kent State Unversty Dept. of Math & Computer Scence LECT-6 Dynamc Programmng 2 Dynamc Programmng Dynamc Programmng, lke the dvde-and-conquer method, solves problems

More information

Maintaining temporal validity of real-time data on non-continuously executing resources

Maintaining temporal validity of real-time data on non-continuously executing resources Mantanng temporal valdty of real-tme data on non-contnuously executng resources Tan Ba, Hong Lu and Juan Yang Hunan Insttute of Scence and Technology, College of Computer Scence, 44, Yueyang, Chna Wuhan

More information

Game Based Virtual Bandwidth Allocation for Virtual Networks in Data Centers

Game Based Virtual Bandwidth Allocation for Virtual Networks in Data Centers Avaable onlne at www.scencedrect.com Proceda Engneerng 23 (20) 780 785 Power Electroncs and Engneerng Applcaton, 20 Game Based Vrtual Bandwdth Allocaton for Vrtual Networks n Data Centers Cu-rong Wang,

More information

A Concurrent Non-Recursive Textured Algorithm for Distributed Multi-Utility State Estimation

A Concurrent Non-Recursive Textured Algorithm for Distributed Multi-Utility State Estimation 1 A Concurrent Non-ecursve Textured Algorthm for Dstrbuted Mult-Utlty State Estmaton Garng M. Huang, Senor Member, IEEE, and Jansheng Le, Student Member, IEEE Abstract: Durng power deregulaton, power companes

More information

Tsinghua University at TAC 2009: Summarizing Multi-documents by Information Distance

Tsinghua University at TAC 2009: Summarizing Multi-documents by Information Distance Tsnghua Unversty at TAC 2009: Summarzng Mult-documents by Informaton Dstance Chong Long, Mnle Huang, Xaoyan Zhu State Key Laboratory of Intellgent Technology and Systems, Tsnghua Natonal Laboratory for

More information

3. CR parameters and Multi-Objective Fitness Function

3. CR parameters and Multi-Objective Fitness Function 3 CR parameters and Mult-objectve Ftness Functon 41 3. CR parameters and Mult-Objectve Ftness Functon 3.1. Introducton Cogntve rados dynamcally confgure the wreless communcaton system, whch takes beneft

More information

CHAPTER 2 PROPOSED IMPROVED PARTICLE SWARM OPTIMIZATION

CHAPTER 2 PROPOSED IMPROVED PARTICLE SWARM OPTIMIZATION 24 CHAPTER 2 PROPOSED IMPROVED PARTICLE SWARM OPTIMIZATION The present chapter proposes an IPSO approach for multprocessor task schedulng problem wth two classfcatons, namely, statc ndependent tasks and

More information

Burst Round Robin as a Proportional-Share Scheduling Algorithm

Burst Round Robin as a Proportional-Share Scheduling Algorithm Burst Round Robn as a Proportonal-Share Schedulng Algorthm Tarek Helmy * Abdelkader Dekdouk ** * College of Computer Scence & Engneerng, Kng Fahd Unversty of Petroleum and Mnerals, Dhahran 31261, Saud

More information

Configuration Management in Multi-Context Reconfigurable Systems for Simultaneous Performance and Power Optimizations*

Configuration Management in Multi-Context Reconfigurable Systems for Simultaneous Performance and Power Optimizations* Confguraton Management n Mult-Context Reconfgurable Systems for Smultaneous Performance and Power Optmzatons* Rafael Maestre, Mlagros Fernandez Departamento de Arqutectura de Computadores y Automátca Unversdad

More information

Efficient Content Distribution in Wireless P2P Networks

Efficient Content Distribution in Wireless P2P Networks Effcent Content Dstrbuton n Wreless P2P Networs Qong Sun, Vctor O. K. L, and Ka-Cheong Leung Department of Electrcal and Electronc Engneerng The Unversty of Hong Kong Pofulam Road, Hong Kong, Chna {oansun,

More information

The Greedy Method. Outline and Reading. Change Money Problem. Greedy Algorithms. Applications of the Greedy Strategy. The Greedy Method Technique

The Greedy Method. Outline and Reading. Change Money Problem. Greedy Algorithms. Applications of the Greedy Strategy. The Greedy Method Technique //00 :0 AM Outlne and Readng The Greedy Method The Greedy Method Technque (secton.) Fractonal Knapsack Problem (secton..) Task Schedulng (secton..) Mnmum Spannng Trees (secton.) Change Money Problem Greedy

More information

Buffering High-Speed Packets with Tri-Stage Memory Array and Its Performance Analysis

Buffering High-Speed Packets with Tri-Stage Memory Array and Its Performance Analysis 000-9825/2005/6(2)28 2005 Journal of Software Vol6, No2,2+, 2,, (,00084) 2 (, 450002) Bufferng Hgh-Speed Packets wth Tr-Stage Memory Array and Its Performance Analyss WANG Peng,2+, YI Peng 2, JIN De-Peng,

More information

Data-Aware Scheduling Strategy for Scientific Workflow Applications in IaaS Cloud Computing

Data-Aware Scheduling Strategy for Scientific Workflow Applications in IaaS Cloud Computing Data-Aware Schedulng Strategy for Scentfc Workflow Applcatons n IaaS Cloud Computng Sd Ahmed Makhlouf*, Belabbas Yagoub LIO Laboratory, Department of Computer Scence, Faculty of Exact and Appled Scences,

More information

Shared Running Buffer Based Proxy Caching of Streaming Sessions

Shared Running Buffer Based Proxy Caching of Streaming Sessions Shared Runnng Buffer Based Proxy Cachng of Streamng Sessons Songqng Chen, Bo Shen, Yong Yan, Sujoy Basu Moble and Meda Systems Laboratory HP Laboratores Palo Alto HPL-23-47 March th, 23* E-mal: sqchen@cs.wm.edu,

More information

Parallel Implementation of Classification Algorithms Based on Cloud Computing Environment

Parallel Implementation of Classification Algorithms Based on Cloud Computing Environment TELKOMNIKA, Vol.10, No.5, September 2012, pp. 1087~1092 e-issn: 2087-278X accredted by DGHE (DIKTI), Decree No: 51/Dkt/Kep/2010 1087 Parallel Implementaton of Classfcaton Algorthms Based on Cloud Computng

More information

Performance Study of Parallel Programming on Cloud Computing Environments Using MapReduce

Performance Study of Parallel Programming on Cloud Computing Environments Using MapReduce Performance Study of Parallel Programmng on Cloud Computng Envronments Usng MapReduce Wen-Chung Shh, Shan-Shyong Tseng Department of Informaton Scence and Applcatons Asa Unversty Tachung, 41354, Tawan

More information

Performance Evaluation of Information Retrieval Systems

Performance Evaluation of Information Retrieval Systems Why System Evaluaton? Performance Evaluaton of Informaton Retreval Systems Many sldes n ths secton are adapted from Prof. Joydeep Ghosh (UT ECE) who n turn adapted them from Prof. Dk Lee (Unv. of Scence

More information

A Dynamic Feedback-based Load Balancing Methodology

A Dynamic Feedback-based Load Balancing Methodology .J. Modern Educaton and Computer Scence, 2017, 12, 57-65 Publshed Onlne December 2017 n MECS (http://www.mecs-press.org/) DO: 10.5815/jmecs.2017.12.07 A Dynamc Feedback-based Load Balancng Methodology

More information

Scheduling Independent Tasks in Heterogeneous Environments under Communication Constraints

Scheduling Independent Tasks in Heterogeneous Environments under Communication Constraints Schedulng Independent Tasks n Heterogeneous Envronments under Communcaton Constrants Petros Lampsas 1 Thanass Loukopoulos 2 Fedon Dmopoulos 1 Marouso Athanasou 1 1 Department of Informatcs and Computer

More information

TripS: Automated Multi-tiered Data Placement in a Geo-distributed Cloud Environment

TripS: Automated Multi-tiered Data Placement in a Geo-distributed Cloud Environment TrpS: Automated Mult-tered Data Placement n a Geo-dstrbuted Cloud Envronment Kwangsung Oh, Abhshek Chandra, and Jon Wessman Department of Computer Scence and Engneerng Unversty of Mnnesota Twn Ctes Mnneapols,

More information

Proper Choice of Data Used for the Estimation of Datum Transformation Parameters

Proper Choice of Data Used for the Estimation of Datum Transformation Parameters Proper Choce of Data Used for the Estmaton of Datum Transformaton Parameters Hakan S. KUTOGLU, Turkey Key words: Coordnate systems; transformaton; estmaton, relablty. SUMMARY Advances n technologes and

More information

Available online at Available online at Advanced in Control Engineering and Information Science

Available online at   Available online at   Advanced in Control Engineering and Information Science Avalable onlne at wwwscencedrectcom Avalable onlne at wwwscencedrectcom Proceda Proceda Engneerng Engneerng 00 (2011) 15000 000 (2011) 1642 1646 Proceda Engneerng wwwelsevercom/locate/proceda Advanced

More information

Optimizing Document Scoring for Query Retrieval

Optimizing Document Scoring for Query Retrieval Optmzng Document Scorng for Query Retreval Brent Ellwen baellwe@cs.stanford.edu Abstract The goal of ths project was to automate the process of tunng a document query engne. Specfcally, I used machne learnng

More information

CS 268: Lecture 8 Router Support for Congestion Control

CS 268: Lecture 8 Router Support for Congestion Control CS 268: Lecture 8 Router Support for Congeston Control Ion Stoca Computer Scence Dvson Department of Electrcal Engneerng and Computer Scences Unversty of Calforna, Berkeley Berkeley, CA 9472-1776 Router

More information

Multi-objective Virtual Machine Placement for Load Balancing

Multi-objective Virtual Machine Placement for Load Balancing Mult-obectve Vrtual Machne Placement for Load Balancng Feng FANG and Bn-Bn Qu,a School of Computer Scence & Technology, Huazhong Unversty Of Scence And Technology, Wuhan, Chna Abstract. The vrtual machne

More information

Towards Autonomous Service Composition in A Grid Environment

Towards Autonomous Service Composition in A Grid Environment Towards Autonomous Servce Composton n A Grd Envronment Wllam K. Cheung +, Jmng Lu +, Kevn H. Tsang +, Raymond K. Wong ++ Department of Computer Scence + Hong Kong Baptst Unversty Hong Kong {wllam,jmng,hhtsang}@comp.hkbu.edu.hk

More information

A Frame Packing Mechanism Using PDO Communication Service within CANopen

A Frame Packing Mechanism Using PDO Communication Service within CANopen 28 A Frame Packng Mechansm Usng PDO Communcaton Servce wthn CANopen Mnkoo Kang and Kejn Park Dvson of Industral & Informaton Systems Engneerng, Ajou Unversty, Suwon, Gyeongg-do, South Korea Summary The

More information

DESIGNING TRANSMISSION SCHEDULES FOR WIRELESS AD HOC NETWORKS TO MAXIMIZE NETWORK THROUGHPUT

DESIGNING TRANSMISSION SCHEDULES FOR WIRELESS AD HOC NETWORKS TO MAXIMIZE NETWORK THROUGHPUT DESIGNING TRANSMISSION SCHEDULES FOR WIRELESS AD HOC NETWORKS TO MAXIMIZE NETWORK THROUGHPUT Bran J. Wolf, Joseph L. Hammond, and Harlan B. Russell Dept. of Electrcal and Computer Engneerng, Clemson Unversty,

More information

A Fast Content-Based Multimedia Retrieval Technique Using Compressed Data

A Fast Content-Based Multimedia Retrieval Technique Using Compressed Data A Fast Content-Based Multmeda Retreval Technque Usng Compressed Data Borko Furht and Pornvt Saksobhavvat NSF Multmeda Laboratory Florda Atlantc Unversty, Boca Raton, Florda 3343 ABSTRACT In ths paper,

More information