The Dual Round Robin Matching Switch with Exhaustive Service

Size: px
Start display at page:

Download "The Dual Round Robin Matching Switch with Exhaustive Service"

Transcription

1 The Dual Round Robin Matching Switch with Exhaustive Sevice Yihan Li, Shivenda S. Panwa, H. Jonathan Chao Abstact Vitual Output Queuing is widely used by fixed-length highspeed switches to ovecome head-of-line blocking. This is done by means of matching algoithms. Maximum matching algoithms have good pefomance, but thei implementation complexity is quite high. Maximal matching algoithms need speedup to guaantee good pefomance. Iteative algoithms (such as PIM and islip) use multiple iteations to convege on a maximal match. The Dual Round-Robin Matching (DRRM) scheme has pefomance simila to islip and lowe implementation complexity. The objective of matching algoithms is to educe the matching ovehead fo each time slot. In this pape we pesent the Exhaustive Sevice Dual Round-Robin Matching (EDRRM) algoithm, which amotizes the cost of a match ove multiple time slots. While EDRRM suffes fom a thoughput below 00% fo small switch sizes, it is conjectued to achieve an asymptotic 00% thoughput unde unifom taffic. Simulations show that it achieves high thoughput unde nonunifom taffic. Its delay pefomance is not sensitive to taffic bustiness, switch size and packet length. In an EDRRM switch cells belonging to the same packet ae tansfeed to the output continuously, which leads to good packet delay pefomance and simplifies the implementation of packet eassembly. Keywods switching, scheduling, Vitual Output Queueing, Dual Round Robin, polling. I. ITRODUCTIO FIXED-LEGTH switching technology is widely accepted as an appoach to achieve high switching efficiency fo high speed packet switches. Vaiable-length IP packets ae segmented into fixed-length cells at inputs and ae eassembled at the outputs. Packet switches based on Input Queuing (IQ) ae desiable fo high speed switching, since the intenal opeation speed is only slightly highe than the input line. Howeve, an Input Queuing switch has a citical dawback [], []: the thoughput is limited to 58:6% due to the head-of-line (HOL) blocking phenomena. Output Queuing (OQ) switches have the optimal delaythoughput pefomance fo all taffic distibutions, but the - times speed-up in the fabic limits the scalability of this achitectue. Vitual Output Queuing (VOQ) is used to ovecome the dawbacks and combine the advantages of an Input Queuing switch and an Output Queuing switch. In a VOQ switch, each input maintains queues, one fo each output. By using VOQ, no additional speedup is equied and HOL blocking can be eliminated. Yihan Li is a Ph.D. candidate in the Electical and Compute Engineeing Depatment, Polytechnic Univesity, Booklyn, Y, USA. yli@photon.poly.edu. Shivenda S. Panwa and H. Jonathan Chao ae on the faculty of the Electical and Compute Engineeing Depatment, Polytechnic Univesity, Booklyn, Y, USA. panwa@catt.poly.edu, chao@antioch.poly.edu. This wok is suppoted in pat by the ew Yok State Cente fo Advanced Technology in Telecommunications (CATT), and also in pat by the ational Science Foundation unde gants ITR00857 and ITR Addess: Electical and Compute Engineeing Depatment, Polytechnic Univesity, Six MetoTech Cente, Booklyn, Y, 0-80, USA. Fax: Consideable wok has been done on scheduling algoithms fo VOQ switches. It has been poved that by using a maximum weight matching algoithm 00% thoughput can be eached fo i.i.d. aivals (unifom o nonunifom) [] [] [5]. But maximum match is not pactical to implement in hadwae due to its complexity, and may not guaantee fainess and quality of sevice. A numbe of pactical maximal matching algoithms have been poposed [6] [7] [8], but maximal matching algoithm cannot achieve as high a thoughput as maximum matching algoithms. Iteative algoithms such as PIM [9] and islip [0] [5], use multiple iteations to convege on a maximal matching. The Dual Round-Robin Matching (DRRM) switch [] [] builds and impoves on the ideas incopoated in islip. It has been poven that DRRM can achieve 00% thoughput unde i.i.d. and unifom taffic []. Futhemoe, the DRRM scheme povides fainess and pevents stavation. It has lowe implementation complexity compaed to algoithms with simila pefomance and is scalable. Accoding to simulation esults [], unde unifom busty taffic, the aveage delay of a DRRM switch vaies appoximately linealy with bust length, but unde nonunifom taffic the thoughput dops below 00%. In this pape a vaiation of DRRM, Exhaustive sevice DRRM (EDRRM) is pesented to impove switching pefomance unde busty and nonunifom taffic. The implementation of EDRRM and DRRM ae compaable with both having lowe complexity than islip. The only pefomance dawback of EDRRM is that it does not achieve 00% thoughput unde unifom taffic fo a ange of switch sizes. We conjectue that fo an EDRRM switch of lage size, thoughputs appoaching 00% ae achievable unde unifom taffic. Simulation esults pesented in this pape and analysis esults in [] suppot, though not igoously pove, this conjectue. Fo smalle switch sizes, we demonstate this popety though simulations. Compaed to DRRM and is- LIP, EDRRM has highe thoughput unde nonunifom taffic. The delay of EDRRM is less sensitive to taffic bustiness, and inceases much slowe with switch size. EDRRM is neithe a maximum matching no a maximal matching algoithm. Unlike any maximum o maximal matching algoithm, which ty to find as many matches as possible in each time slot, EDRRM achieves efficiency by looking at the matching ovehead ove time. In EDRRM the cost in wasted slot times to get a match may be lage, but the cost is amotized ove a VOQ busy peiod. We believe that this is a new appoach with both theoetical and pactical implications. Most of the pevious wok only consides the cell delay that a cell suffes fom the time it entes a VOQ to the time it is tansfeed to the destination output pot. Additional delay is incued at the Output Reassembly Module (ORM) of each output to eassemble packets []. Multiple queues ae needed at each ORM if cells belonging to diffeent packets ae inteleaved at the

2 same output. When a cell it tansfeed though the switch fabic to the output, it is deliveed into one of the queues of the ORM. The cells belonging to the same packet will be deliveed to the same queue and can only leave the queue until the whole packet is eassembled. The total delay a packet suffes includes the cell delay and the time needed fo eassembly. Thus the cell delay is not enough to evaluate the delay pefomance of a packet switch. In this pape we measue the packet delay as well as the cell delay of EDRRM and compae them to those of DRRM and islip. It shows that unde unifom i.i.d. taffic, even when the cell delay of EDRRM is highe, the packet delay is lowe, and it is not sensitive to switch size and packet length. At the same time, since all the cells belonging to the same packet ae tansfeed to the output continuously, only one queue is needed in each ORM, which futhe simplifies the switch implementation. In section II we biefly oveview the DRRM algoithm and its pefomance. In section III the EDRRM algoithm is descibed in detail. In section IV the thoughput, aveage cell delay and aveage packet delay pefomance of an EDRRM switch is discussed. II. THE DRRM ALGORITHM AD ITS PERFORMACE In the DRRM scheme each input pot maintains VOQs. The algoithm has two steps: Step : Request. Each input sends an output equest coesponding to the fist nonempty VOQ in a fixed ound-obin ode, stating fom the cuent position of the pointe. The pointe emains at that nonempty VOQ if the selected output is not ganted in step. The pointe of the input abite is incemented by one location beyond the selected output if, and only if, the equest is ganted in step. Step : Gant. If an output eceives one o moe equests, it chooses the one that appeas next in a fixed ound-obin schedule stating fom the cuent position of the pointe. The output notifies each equesting input whethe o not its equest was ganted. The pointe of the output abite is incemented to one location beyond the ganted input. If thee ae no equests, the pointe emains whee it is. The pefomance of the DRRM scheme has been shown and compaed with that of islip in []. Unde unifom and i.i.d. taffic the thoughput of a DRRM switch is 00%, and the aveage cell delay inceases with switch size fo a given. The pefomance unde nonunifom taffic is also consideed in []. Simulation esults show that unde the hot-spot taffic, thoughput fo the hot-spot output of a DRRM switch is 00%. Fo anothe nonunifom taffic scenaio simulations show that the thoughput of both DRRM and islip dops below 00%. III. THE EXHAUSTIVE SERVICE DRRM SCHEME In the DRRM scheme, when an input and an output ae matched, only one cell will be tansfeed fom the input to the matched output. Afte that both the input and the output will incement thei pointes by one and in the next time slot this input-output pai will have the lowest pioity to get matched to each othe. This behavio is simila to the limited sevice policy [5] in a polling system. In ode to impove on DRRM s pefomance unde non-unifom taffic, we modified the DRRM scheme so that wheneve an input is matched to an output, all the cells in the coesponding VOQ will be tansfeed in the following time slots befoe any othe VOQ of the same input can be seved. This is called the exhaustive sevice policy [5] in polling systems. We theefoe call this the Exhaustive sevice DRRM (EDRRM) scheme. In EDRRM, the pointes of inputs and outputs ae updated in a diffeent way fom DRRM. In a time slot if an input is matched to an output, one cell in the coesponding VOQ will be tansfeed. Afte that, if the VOQ becomes empty, the input will update its abite pointe to the next location in a fixed ode; othewise, the pointe will emain at the cuent VOQ so that a equest will be sent to the same output in the next time slot. If an input sends a equest to an output but gets no gant, the input will update its abite pointe to the next location in a fixed ode, which is diffeent fom DRRM whee the input pointe will emain whee it is until it gets a gant. The eason fo this modification is as follows. In EDRRM if an input cannot get a gant fom an output, it means that the output is most likely in a stable maiage with anothe input fo all the cells waiting in the VOQ, and the unsuccessful input is likely to wait fo a long time to get a gant fom this output. It will be bette fo the input to seach fo anothe fee output than to wait fo this busy one. Since an output has no idea if the cuently seved VOQ will become empty afte this sevice, outputs will not update thei abite pointes afte cell tansfe. A detailed desciption of the two step EDRRM algoithm follows: Step : Request. Each input moves its pointe to the fist nonempty VOQ in a fixed ound-obin ode, stating fom the cuent position of the pointe, and sends a equest to the output coesponding to the VOQ. The pointe of the input abite is incemented by one location beyond the selected output if the equest is not ganted in Step, o if the equest is ganted and afte one cell is seved this VOQ becomes empty. Othewise, the pointe emains at that (nonempty) VOQ. Step : Gant. If an output eceives one o moe equests, it chooses the one that appeas next in a fixed ound-obin schedule stating fom the cuent position of the pointe. The output notifies each equesting input whethe o not its equest was ganted. The pointe of the output abite emains at the ganted input. If thee ae no equests, the pointe emains whee it is. Input Input Input Input Input Abites Request Phase Gant Phase g equest Fig.. An example of EDRRM gant g Output Abites Figue shows an example of the EDRRM abitation algoithm.,, and ae abite pointes fo inputs,, and g g

3 , and g, g, g and g ae abite pointes fo outputs,, and. At the beginning of the time slot points to output while g does not point to input, which means that in the last time slot input was not matched to output, and now input equests output fo a new sevice. Similaly, equests output fo a new sevice. Since points to output and g points to input, it is possible that in the last time slot input was matched to output and in this time slot output will continue to tansfe a cell fom input because the VOQ is not empty. Input and output have a simila situation with input and output. In the gant phase, output gants the only equest it eceives fom input and updates g to, output gants the equest fom input and output gants the equest fom input. The equest fom input to output is not ganted, so moves to. By the end of this time slot, the st VOQ of input and the d VOQ of input ae still nonempty so that and ae not updated; is updated to because the nd VOQ of input becomes empty. The implementation complexity of EDRRM s switching fabic is identical to that of DRRM. Since the opeational step and data exchange is limited, the DRRM abitation mechanism can be implemented in a distibuted manne to make the switch simple and moe scalable. The length of each contol message in DRRM is only th of that in islip. In [] it is shown that by using a token-tunneling technique a switch capacity of moe than one teabit/sec is achievable with existing electonic technology. The ORM of EDRRM is simple than that of DRRM since only one queue is maintained. Usually when cells of diffeent packets ae inteleaved, queues ae needed in each ORM, one fo each input. In the next section, we will show that EDRRM has pefomance compaable with DRRM and islip unde unifom independent taffic, and has bette pefomance unde busty taffic and nonunifom taffic. IV. SIMULATED PERFORMACE OF EDRRM In this section the thoughput and delay pefomance of EDRRM is shown by simulation esults and compaed with that of DRRM. A. Thoughput DRRM is poved to have 00% thoughput unde unifom and i.i.d. taffic in []. Accoding to simulation esults the thoughput of EDRRM unde unifom and i.i.d. taffic is close to, but not quite 00%. Figue shows the thoughput of EDRRM with diffeent switch size unde unifom and i.i.d. taffic with 00% aival ate. ote that the thoughput fist deceased and then inceases with switch size. We conjectue that fo lage the thoughput will appoach 00% asymptotically. This conjectue is futhe suppoted by the analysis in []. Long simulation un lengths pevented us fom consideing switch sizes lage than. While this is cetainly a weakness of EDRRM as compaed to DRRM and islip, we believe that this is an acceptable tadeoff given its pefomance advantages unde moe typical taffic ings. We will conside these next. In this pape we conside fou nonunifom taffic pattens and compae the thoughput of EDRRM to those of DRRM and is- LIP. thoughput % conf. int switch size Fig.. The thoughput of EDRRM switch unde unifom taffic Patten : Hot-spot taffic, which efes to a taffic patten whee many inputs send taffic to one output line (the hot-spot) at the same time, thus oveing it. The hot-spot thoughput of EDRRM is 00%, which is the same as that of DRRM and highe than that of islip. Patten : In this patten the aival ate fo each input is the same, while the ing fo each output is also the same. Fo input i a faction p,» p<of aivals ae destined to output i, and othe aivals ae unifomly destined to othe outputs. When p =, this coesponds to the unifom case. When p =, all the aiving cells of input i ae destined to output i. Figue compaes the thoughput of EDRRM, DRRM and is- LIP unde this taffic patten fo diffeent switch sizes. It shows that EDRRM has highe thoughput than DRRM. The figue also shows that fo EDRRM switch, the thoughput unde unifom aivals (when p = ) is compaable to the thoughput unde nonunifom aivals (when p> ). In a DRRM switch each VOQ gets unifom sevice. On the othe hand, in an EDRRM switch all the cells in a VOQ will be seved when the VOQ wins the sevice no matte what the aival ate is fo this VOQ. When the system is stable, the sevice ate fo a VOQ is close to the aival ate, which leads to a high thoughput. ote that anothe attactive featue of EDRRM is that it automatically adapts to changes in taffic flow though the switch, i.e., no paamete has to be set to optimize pefomance fo a given non-unifom taffic flow. This is impotant since it simplifies switch taffic management as taffic flows change with time. Also note that, as in Figue, the oveall thoughput fist dops with switch size and then inceases with. We see that with inceasing switch size, the EDRRM switch can maintain close to 00% thoughput fo this nonunifom taffic patten. Patten : In this patten the aival ate fo each input is the same. Fo input i a faction f of aivals ae destined to output i, and othe aivals ae destined to output (i +)mod. Table I compaes the thoughput of EDRRM, DRRM and islip unde this taffic patten fo diffeent f with switch size. EDRRM has a highe thoughput than DRRM o islip. Patten : In this patten the aival ate fo each input is the

4 TABLE I THE THROUGHPUT UDER OUIFORM TRAFFIC PATTER f EDRRM DRRM islip TABLE II THE THROUGHPUT UDER OUIFORM TRAFFIC PATTER size 8 6 EDRRM DRRM islip same. Fom input i the taffic to output i + j is two times the to output i + j +, 0» j». Table II compaes the thoughput of EDRRM, DRRM and islip unde this taffic patten fo diffeent switch sizes. The thoughput of EDRRM is always above 90%, which is much highe than the thoughputs of DRRM and islip. Though not pesented hee, a limit on the maximum numbe of cells that can be seved continuously in a VOQ can be enfoced by means of a counte to avoid unfainess in case one output is occupied fo long peiods of time by one input unde some exteme taffic patten EDRRM x EDRRM 8x8 EDRRM 6x6 EDRRM x DRRM x DRRM 8x8 DRRM 6x6 DRRM x islip x islip 8x8 islip 6x6 islip x aveage cell delay of an EDRRM switch unde unifom taffic and a heavy is lage than that of a DRRM switch. This is due to the moe efficient DRRM scheduling mechanism unde unifom, heavy taffic. mean cell delay (time slots) EDRRM x EDRRM 8x8 EDRRM 6x6 DRRM x DRRM 8x8 DRRM 6x Fig.. The aveage cell delay unde unifom taffic fo diffeent switch sizes Figue 5 shows the aveage cell delay of an EDRRM switch, a DRRM switch and an islip switch with switch size 6 6 unde unifom busty taffic with a geometically distibuted bust length of 6 cells. It shows that the aveage delays of DRRM and islip ae simila to each othe and much lage than the aveage delay of EDRRM unde heavy. Since the pefomance of islip and DRRM ae oughly compaable [], we will compae the pefomance of EDRRM with DRRM fom this point onwad EDRRM DRRM islip thoughput nonunifom paamete p Fig.. The thoughput fo diffeent switch sizes unde nonunifom taffic patten B. Aveage cell delay Figue shows the aveage cell delay of EDRRM and DRRM unde unifom and i.i.d. taffic with diffeent switch sizes. The Fig. 5. The aveage cell delay of a 6x6 switch unde unifom busty taffic with bust length = 6 cells Figue 6 compaes the aveage cell delay of an EDRRM switch and a DRRM switch with switch size 6 6 unde unifom and busty taffic with diffeent bust lengths. It shows that the aveage delay of a DRRM switch inceases appoximately linealy with bust length, which is simila to the behavio of

5 5 an EDRRM switch unde light. Significantly, unde heavy the aveage delay of an EDRRM switch does not change much with diffeent bust lengths and is much smalle than that of a DRRM switch fo long bust lengths. mean cell delay (time slots) EDRRM, bust=0 cells EDRRM, bust=6 cells EDRRM, bust= cells DRRM, bust=0 cells DRRM, bust=6 cells DRRM, bust= cells Fig. 6. The aveage cell delay unde unifom and busty taffic with diffeent bust lengths Figue 7 compaes the aveage cell delays of EDRRM and DRRM switches with diffeent switch sizes when the bust length is 6. We can see that as the switch size inceases the aveage cell delay of a DRRM switch gows apidly, while the aveage cell delay of an EDRRM switch gows moe slowly. This indicates that cell delay fo EDRRM is less sensitive to switch size than DRRM fo busty taffic. mean cell delay (time slots) EDRRM x EDRRM 8x8 EDRRM 6x6 EDRRM x DRRM x DRRM 8x8 DRRM 6x6 DRRM x Fig. 7. The aveage cell delay unde unifom and busty taffic fo diffeent switch sizes The eason why EDRRM is not sensitive to busty taffic is that all the cells aive within the same bust will be seved continuously so that little futhe delay will be suffeed because of the bustiness. In DRRM o islip, the sevice of cells in a bust is inteleaved with sevice to othe VOQ s, which may incu additional ovehead due to unsuccessful equests. Fo cells belonging to a bust at a VOQ, when one cell is seved, the next one cannot get sevice befoe all othe nonempty VOQs at the same input have been seved once. C. Aveage packet delay DRRM and EDRRM ae fixed-length switching algoithms. Vaiable-length IP packets ae segmented into fixed-length cells at the inputs, and the cells will be placed in the coesponding VOQ one by one. When a cell is tansfeed to its destination output, it will stay in a buffe and wait fo the othe cells in the same packet. Afte the complete eception of all the cells coming fom the same packet, these cells will be eassembled into a packet. The delay a cell suffeed befoe it is eassembled into a packet and deliveed to its destination includes the cell delay discussed in the last subsection and the waiting time at the output eassembly buffe. So the cell delay pefomance is not sufficient to evaluate the packet delay pefomance of a fixedlength switch. In this subsection, we conside aveage packet delay pefomance fo the DRRM and EDRRM switches. Afte a packet is segmented into cells, one cell will be put into the VOQ in each time slot. As in [], the packet delay of a packet is measued fom the time when the last cell of the packet gets into the VOQ until the time when the same last cell is tansfeed to its destined output EDRRM DRRM islip Fig. 8. The aveage packet delay when the packet size is 0 cells Figue 8 compaes the aveage packet delays of EDRRM, DRRM and islip when the switch size is 6x6 and the packet size is 0 cells. It shows that the packet delays of DRRM and islip ae simila to each othe and lage than that of EDRRM. Figue 9 shows the aveage packet delay of EDRRM and DRRM fo diffeent switch sizes unde unifom taffic with packet size 0. We can see that the aveage packet delay of EDRRM is always compaable with o smalle than that of DRRM when the switch size is lage than, even when the aveage cell delay of DRRM is smalle than that of EDRRM. The eason is that in DRRM, when one cell in a VOQ is seved, the next cell in the same VOQ needs to wait fo its next tun. Fo

6 EDRRM, x EDRRM, 8x8 EDRRM, 6x6 EDRRM, x DRRM, x DRRM, 8x8 DRRM, 6x6 DRRM, x Fig. 9. The aveage packet delay of EDRRM and DRRM unde unifom taffic with diffeent switch sizes example, unde heavy when all the VOQs ae nonempty, afte a cell is tansfeed the next cell in the same VOQ needs to wait at least time slots to get a sevice. Fo the cells in one packet, each cell needs to wait some time in the output eassembly buffe fo the next cell except the last one. On the othe hand, in EDRRM all the cells in the same packet will be seved continuously, and the numbe of time slots that the packet needs fo eassemble in an output buffe is the packet size in cells. Theefoe the packet delay of an EDRRM switch is the cell delay fo the fist cell of the packet plus the packet size in cell slot times. Indeed, an EDRRM switch is eally a packet switch and not a cell switch emulating a packet switch since packets ae sent contiguously though the switch fabic. As a esult, EDRRM does not need a complex eassembly buffe at each output pot EDRRM, packet size=5 EDRRM, packet size=0 EDRRM, packet size=0 DRRM, packet size=5 DRRM, packet size=0 DRRM, packet size= Fig. 0. The aveage packet delay when the packet size is 0 cells Figue 0 shows the aveage packet delays of a 6x6 EDRRM switch and a DRRM switch with diffeent packet sizes. We can see that fo an EDRRM switch, unde light, longe packets suffe longe packet delay, while unde heavy the aveage packet delays fo packets with diffeent size ae simila. On the othe hand, in a DRRM switch the aveage packet delay inceases linealy with the packet size. Though we do not pesent it hee, delay pefomance fo pioity taffic (e.g. eal-time taffic) can be enhanced by intoducing a high and low pioity queue in each VOQ. V. COCLUSIOS The EDRRM algoithm is a vaiation of the DRRM scheduling algoithm. The implementation complexity of EDRRM s switching fabic is the same as that of DRRM, while packet eassembly is simple than most othe popula matching schemes. In an EDRRM switch when an input is matched with an output all the cells in the coesponding VOQ ae seved continuously befoe any othe VOQ of the same input can be seved. The aveage cell delay of an EDRRM switch can be analyzed by using an exhaustive andom polling system model []. The pefomance of an EDRRM switch is compaable to o bette than a DRRM switch o an islip switch fo most taffic scenaios. The only exception is the maximum thoughput unde unifom taffic. Unde unifom i.i.d. taffic, an EDRRM switch has a lage aveage cell delay than a DRRM switch, but its aveage packet delay is lowe and not sensitive to eithe switch size o packet size. Futhemoe, EDRRM is not sensitive to taffic bustiness. Unde nonunifom taffic the thoughputs of a DRRM switch and an islip switch dop well below 00%, while the thoughput of an EDRRM switch is close to 00%. REFERECES [] M. J. Kaol, M. Hluchyj, and S. Mogan, Input vs. output queuing on a space-division packet switch, Poc. GLOBECOM 986, pp [] M. J. Kaol, M. Hluchyj, and S. Mogan, Input vesus output queuing on a space-division packet switch, IEEE Tans. on Communications, vol.5, pp. 7-56, 987. []. McKeown, V. Ananthaam, and J. Waland, Achieving 00% thoughput in an input-queued switch, IEEE IFOCOM 96, pp []. McKeown, A. Mekkittikul, V. Ananthaam and J. Waland, Achieving 00% thoughput in an Input-Queued switch, IEEE Tans. Communications, vol. 7, o. 8, pp , Aug [5]. McKeown, Scheduling algoithms fo input-queued cell switches, Ph.D. Thesis, UC Bekeley, May 995. [6] A. Chany, P. Kishna,. Patel and R. Simcoe, Algoithms fo poviding bandwidth and delay guaantees in Input-Buffeed cossbas with speedup, IWQOS 98, May 998. [7] P. Kishna,. S. Patel, A.Chany and R. Simcoe, On the speedup equied fo wok-conseving cossba switches, IWQOS 98, May 998. [8] A. Mekkittikul and. McKeown, A pactical scheduling algoithm to achieve 00% thoughput in input-queued switches, IEEE IFOCOM 98, Vol, pp , Apil 998. [9] T. E. Andeson, S. S. Owicki, J. B. Saxe and C. P. Thacke, High speed switch scheduling fo local aea netwoks, ACM Tans. on Compute Systems, vol., o., pp. 9-5, ov. 99. [0]. McKeown, The islip scheduling algoithm fo Input-Queued switches, IEEE/ACM Tans. etwoking, vol. 7, pp. 88-0, Apil 999. [] H. J. Chao, Satun: a teabit packet switch using Dual Round-Robin, IEEE Communication Magazine, vol. 8, pp. 78-8, Dec [] Y. Li, S. Panwa, H. J. Chao, On the pefomance of a Dual Round-Robin switch, IEEE IFOCOM 00, vol., pp , Apil 00. [] Y. Li, S. Panwa, H. J. Chao, Pefomance analysis of an Exhaustive Sevice Dual Round-Robin scheduling slgoithm, CATT Technical Repot, ov. 00. [] Y. A. Masan, A. Bianco, P. Giaccone, E. Leonadi, F. ei, Packet Scheduling in Input-Queued Cell-Based Switches, IEEE IFOCOM 00, vol., pp , Apil 00. [5] H. Takagi, Queueing anallysis of polling models: an update, Stochastic Analysis of Compute and Communication Systems, ew Yok: Elsevie Science and B. V. oth Holland, pp

F cepted as an approach to achieve high switching efficiency

F cepted as an approach to achieve high switching efficiency The Dual Round Robin Matching Switch with Exhaustive Service Yihan Li, Shivendra Panwar, H. Jonathan Chao AbsrmcrVirtual Output Queuing is widely used by fixedlength highspeed switches to overcome headofline

More information

IP Network Design by Modified Branch Exchange Method

IP Network Design by Modified Branch Exchange Method Received: June 7, 207 98 IP Netwok Design by Modified Banch Method Kaiat Jaoenat Natchamol Sichumoenattana 2* Faculty of Engineeing at Kamphaeng Saen, Kasetsat Univesity, Thailand 2 Faculty of Management

More information

Slotted Random Access Protocol with Dynamic Transmission Probability Control in CDMA System

Slotted Random Access Protocol with Dynamic Transmission Probability Control in CDMA System Slotted Random Access Potocol with Dynamic Tansmission Pobability Contol in CDMA System Intaek Lim 1 1 Depatment of Embedded Softwae, Busan Univesity of Foeign Studies, itlim@bufs.ac.k Abstact In packet

More information

Worst-Case Delay Bounds for Uniform Load-Balanced Switch Fabrics

Worst-Case Delay Bounds for Uniform Load-Balanced Switch Fabrics Wost-Case Delay Bounds fo Unifom Load-Balanced Switch Fabics Spyidon Antonakopoulos, Steven Fotune, Rae McLellan, Lisa Zhang Bell Laboatoies, 600 Mountain Ave, Muay Hill, NJ 07974 fistname.lastname@alcatel-lucent.com

More information

ANALYTIC PERFORMANCE MODELS FOR SINGLE CLASS AND MULTIPLE CLASS MULTITHREADED SOFTWARE SERVERS

ANALYTIC PERFORMANCE MODELS FOR SINGLE CLASS AND MULTIPLE CLASS MULTITHREADED SOFTWARE SERVERS ANALYTIC PERFORMANCE MODELS FOR SINGLE CLASS AND MULTIPLE CLASS MULTITHREADED SOFTWARE SERVERS Daniel A Menascé Mohamed N Bennani Dept of Compute Science Oacle, Inc Geoge Mason Univesity 1211 SW Fifth

More information

HISTOGRAMS are an important statistic reflecting the

HISTOGRAMS are an important statistic reflecting the JOURNAL OF L A T E X CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 1 D 2 HistoSketch: Disciminative and Dynamic Similaity-Peseving Sketching of Steaming Histogams Dingqi Yang, Bin Li, Laua Rettig, and Philippe

More information

Journal of World s Electrical Engineering and Technology J. World. Elect. Eng. Tech. 1(1): 12-16, 2012

Journal of World s Electrical Engineering and Technology J. World. Elect. Eng. Tech. 1(1): 12-16, 2012 2011, Scienceline Publication www.science-line.com Jounal of Wold s Electical Engineeing and Technology J. Wold. Elect. Eng. Tech. 1(1): 12-16, 2012 JWEET An Efficient Algoithm fo Lip Segmentation in Colo

More information

Adaptation of TDMA Parameters Based on Network Conditions

Adaptation of TDMA Parameters Based on Network Conditions Adaptation of TDMA Paametes Based on Netwok Conditions Boa Kaaoglu Dept. of Elect. and Compute Eng. Univesity of Rocheste Rocheste, NY 14627 Email: kaaoglu@ece.ocheste.edu Tolga Numanoglu Dept. of Elect.

More information

A Two-stage and Parameter-free Binarization Method for Degraded Document Images

A Two-stage and Parameter-free Binarization Method for Degraded Document Images A Two-stage and Paamete-fee Binaization Method fo Degaded Document Images Yung-Hsiang Chiu 1, Kuo-Liang Chung 1, Yong-Huai Huang 2, Wei-Ning Yang 3, Chi-Huang Liao 4 1 Depatment of Compute Science and

More information

On Adaptive Bandwidth Sharing with Rate Guarantees

On Adaptive Bandwidth Sharing with Rate Guarantees On Adaptive Bandwidth Shaing with Rate Guaantees N.G. Duffield y T. V. Lakshman D. Stiliadis y AT&T Laboatoies Bell Labs Rm A175, 180 Pak Avenue Lucent Technologies Floham Pak, 101 Cawfods Cone Road NJ

More information

Modeling a shared medium access node with QoS distinction

Modeling a shared medium access node with QoS distinction Modeling a shaed medium access node with QoS distinction Matthias Gies, Jonas Geutet Compute Engineeing and Netwoks Laboatoy (TIK) Swiss Fedeal Institute of Technology Züich CH-8092 Züich, Switzeland email:

More information

Detection and Recognition of Alert Traffic Signs

Detection and Recognition of Alert Traffic Signs Detection and Recognition of Alet Taffic Signs Chia-Hsiung Chen, Macus Chen, and Tianshi Gao 1 Stanfod Univesity Stanfod, CA 9305 {echchen, macuscc, tianshig}@stanfod.edu Abstact Taffic signs povide dives

More information

Modelling, simulation, and performance analysis of a CAN FD system with SAE benchmark based message set

Modelling, simulation, and performance analysis of a CAN FD system with SAE benchmark based message set Modelling, simulation, and pefomance analysis of a CAN FD system with SAE benchmak based message set Mahmut Tenuh, Panagiotis Oikonomidis, Peiklis Chachalakis, Elias Stipidis Mugla S. K. Univesity, TR;

More information

Controlled Information Maximization for SOM Knowledge Induced Learning

Controlled Information Maximization for SOM Knowledge Induced Learning 3 Int'l Conf. Atificial Intelligence ICAI'5 Contolled Infomation Maximization fo SOM Knowledge Induced Leaning Ryotao Kamimua IT Education Cente and Gaduate School of Science and Technology, Tokai Univeisity

More information

EE 6900: Interconnection Networks for HPC Systems Fall 2016

EE 6900: Interconnection Networks for HPC Systems Fall 2016 EE 6900: Inteconnection Netwoks fo HPC Systems Fall 2016 Avinash Kaanth Kodi School of Electical Engineeing and Compute Science Ohio Univesity Athens, OH 45701 Email: kodi@ohio.edu 1 Acknowledgement: Inteconnection

More information

Dynamic Multiple Parity (DMP) Disk Array for Serial Transaction Processing

Dynamic Multiple Parity (DMP) Disk Array for Serial Transaction Processing IEEE TRANSACTIONS ON COMPUTERS, VOL. 50, NO. 9, SEPTEMBER 200 949 Dynamic Multiple Paity (DMP) Disk Aay fo Seial Tansaction Pocessing K.H. Yeung, Membe, IEEE, and T.S. Yum, Senio Membe, IEEE AbstactÐThe

More information

A Memory Efficient Array Architecture for Real-Time Motion Estimation

A Memory Efficient Array Architecture for Real-Time Motion Estimation A Memoy Efficient Aay Achitectue fo Real-Time Motion Estimation Vasily G. Moshnyaga and Keikichi Tamau Depatment of Electonics & Communication, Kyoto Univesity Sakyo-ku, Yoshida-Honmachi, Kyoto 66-1, JAPAN

More information

AN ANALYSIS OF COORDINATED AND NON-COORDINATED MEDIUM ACCESS CONTROL PROTOCOLS UNDER CHANNEL NOISE

AN ANALYSIS OF COORDINATED AND NON-COORDINATED MEDIUM ACCESS CONTROL PROTOCOLS UNDER CHANNEL NOISE AN ANALYSIS OF COORDINATED AND NON-COORDINATED MEDIUM ACCESS CONTROL PROTOCOLS UNDER CHANNEL NOISE Tolga Numanoglu, Bulent Tavli, and Wendi Heinzelman Depatment of Electical and Compute Engineeing Univesity

More information

Ethernet PON (epon): Design and Analysis of an Optical Access Network.

Ethernet PON (epon): Design and Analysis of an Optical Access Network. Ethenet PON epon: Design and Analysis of an Optical Access Netwo. Glen Kame Depatment of Compute Science Univesity of Califonia, Davis, CA 9566, USA Tel:.530.297.527; Fax:.530.297.530 E-mail: ame@cs.ucdavis.edu

More information

Conversion Functions for Symmetric Key Ciphers

Conversion Functions for Symmetric Key Ciphers Jounal of Infomation Assuance and Secuity 2 (2006) 41 50 Convesion Functions fo Symmetic Key Ciphes Deba L. Cook and Angelos D. Keomytis Depatment of Compute Science Columbia Univesity, mail code 0401

More information

A Stable Traffic Engineering Technique for Performance Enhancement of the Non-TCP Elastic Flows *

A Stable Traffic Engineering Technique for Performance Enhancement of the Non-TCP Elastic Flows * JOURNAL OF INFORMATION SCIENCE AND ENGINEERING 30, 1115-1129 (2014) A Stable Taffic Engineeing Technique fo Pefomance Enhancement of the Non-TCP Elastic Flows * Ian Telecom Reseach Cente (ITRC) Tehan,

More information

An Improved Resource Reservation Protocol

An Improved Resource Reservation Protocol Jounal of Compute Science 3 (8: 658-665, 2007 SSN 549-3636 2007 Science Publications An mpoved Resouce Resevation Potocol Desie Oulai, Steven Chambeland and Samuel Piee Depatment of Compute Engineeing

More information

WIRELESS sensor networks (WSNs), which are capable

WIRELESS sensor networks (WSNs), which are capable IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS, VOL. XX, NO. XX, XXX 214 1 Lifetime and Enegy Hole Evolution Analysis in Data-Gatheing Wieless Senso Netwoks Ju Ren, Student Membe, IEEE, Yaoxue Zhang, Kuan

More information

Performance Optimization in Structured Wireless Sensor Networks

Performance Optimization in Structured Wireless Sensor Networks 5 The Intenational Aab Jounal of Infomation Technology, Vol. 6, o. 5, ovembe 9 Pefomance Optimization in Stuctued Wieless Senso etwoks Amine Moussa and Hoda Maalouf Compute Science Depatment, ote Dame

More information

Simulation and Performance Evaluation of Network on Chip Architectures and Algorithms using CINSIM

Simulation and Performance Evaluation of Network on Chip Architectures and Algorithms using CINSIM J. Basic. Appl. Sci. Res., 1(10)1594-1602, 2011 2011, TextRoad Publication ISSN 2090-424X Jounal of Basic and Applied Scientific Reseach www.textoad.com Simulation and Pefomance Evaluation of Netwok on

More information

On using circuit-switched networks for file transfers

On using circuit-switched networks for file transfers On using cicuit-switched netwoks fo file tansfes Xiuduan Fang, Malathi Veeaaghavan Univesity of Viginia Email: {xf4c, mv5g}@viginia.edu Abstact High-speed optical cicuit-switched netwoks ae being deployed

More information

Prioritized Traffic Recovery over GMPLS Networks

Prioritized Traffic Recovery over GMPLS Networks Pioitized Taffic Recovey ove GMPLS Netwoks 2005 IEEE. Pesonal use of this mateial is pemitted. Pemission fom IEEE mu be obtained fo all othe uses in any cuent o futue media including epinting/epublishing

More information

THE THETA BLOCKCHAIN

THE THETA BLOCKCHAIN THE THETA BLOCKCHAIN Theta is a decentalized video steaming netwok, poweed by a new blockchain and token. By Theta Labs, Inc. Last Updated: Nov 21, 2017 esion 1.0 1 OUTLINE Motivation Reputation Dependent

More information

Communication vs Distributed Computation: an alternative trade-off curve

Communication vs Distributed Computation: an alternative trade-off curve Communication vs Distibuted Computation: an altenative tade-off cuve Yahya H. Ezzeldin, Mohammed amoose, Chistina Fagouli Univesity of Califonia, Los Angeles, CA 90095, USA, Email: {yahya.ezzeldin, mkamoose,

More information

arxiv: v2 [physics.soc-ph] 30 Nov 2016

arxiv: v2 [physics.soc-ph] 30 Nov 2016 Tanspotation dynamics on coupled netwoks with limited bandwidth Ming Li 1,*, Mao-Bin Hu 1, and Bing-Hong Wang 2, axiv:1607.05382v2 [physics.soc-ph] 30 Nov 2016 1 School of Engineeing Science, Univesity

More information

Hierarchically Clustered P2P Streaming System

Hierarchically Clustered P2P Streaming System Hieachically Clusteed P2P Steaming System Chao Liang, Yang Guo, and Yong Liu Polytechnic Univesity Thomson Lab Booklyn, NY 11201 Pinceton, NJ 08540 Abstact Pee-to-pee video steaming has been gaining populaity.

More information

RT-WLAN: A Soft Real-Time Extension to the ORiNOCO Linux Device Driver

RT-WLAN: A Soft Real-Time Extension to the ORiNOCO Linux Device Driver 1 RT-WLAN: A Soft Real-Time Extension to the ORiNOCO Linux Device Dive Amit Jain Daji Qiao Kang G. Shin The Univesity of Michigan Ann Abo, MI 4819, USA {amitj,dqiao,kgshin@eecs.umich.edu Abstact The cuent

More information

Optical Flow for Large Motion Using Gradient Technique

Optical Flow for Large Motion Using Gradient Technique SERBIAN JOURNAL OF ELECTRICAL ENGINEERING Vol. 3, No. 1, June 2006, 103-113 Optical Flow fo Lage Motion Using Gadient Technique Md. Moshaof Hossain Sake 1, Kamal Bechkoum 2, K.K. Islam 1 Abstact: In this

More information

GTOC 9, Multiple Space Debris Rendezvous Trajectory Design in the J2 environment

GTOC 9, Multiple Space Debris Rendezvous Trajectory Design in the J2 environment GTOC 9, Multiple Space Debis Rendezvous Tajectoy Design in the J envionment Macus Hallmann, Makus Schlottee, Ansga Heidecke, Maco Sagliano Fedeico Fumenti, Volke Maiwald, René Schwaz Institute of Space

More information

Combinatorial Mobile IP: A New Efficient Mobility Management Using Minimized Paging and Local Registration in Mobile IP Environments

Combinatorial Mobile IP: A New Efficient Mobility Management Using Minimized Paging and Local Registration in Mobile IP Environments Wieless Netwoks 0, 3 32, 200 200 Kluwe Academic Publishes. Manufactued in The Nethelands. Combinatoial Mobile IP: A New Efficient Mobility Management Using Minimized Paging and Local Registation in Mobile

More information

Quality Aware Privacy Protection for Location-based Services

Quality Aware Privacy Protection for Location-based Services In Poceedings of the th Intenational Confeence on Database Systems fo Advanced Applications (DASFAA 007), Bangkok, Thailand, Apil 9-, 007. Quality Awae Pivacy Potection fo Location-based Sevices Zhen Xiao,,

More information

CLOUD based storage systems are emerging to gain significant

CLOUD based storage systems are emerging to gain significant IEEE TRANSACTIONS ON CLOUD COMPUTING, VOL. 5, NO. 2, APRIL-JUNE 2017 221 On the Latency and Enegy Efficiency of Distibuted Stoage Systems Akshay Kuma, Student Membe, IEEE, Ravi Tandon, Membe, IEEE, and

More information

An Unsupervised Segmentation Framework For Texture Image Queries

An Unsupervised Segmentation Framework For Texture Image Queries An Unsupevised Segmentation Famewok Fo Textue Image Queies Shu-Ching Chen Distibuted Multimedia Infomation System Laboatoy School of Compute Science Floida Intenational Univesity Miami, FL 33199, USA chens@cs.fiu.edu

More information

Towards Adaptive Information Merging Using Selected XML Fragments

Towards Adaptive Information Merging Using Selected XML Fragments Towads Adaptive Infomation Meging Using Selected XML Fagments Ho-Lam Lau and Wilfed Ng Depatment of Compute Science and Engineeing, The Hong Kong Univesity of Science and Technology, Hong Kong {lauhl,

More information

Dynamic Topology Control to Reduce Interference in MANETs

Dynamic Topology Control to Reduce Interference in MANETs Dynamic Topology Contol to Reduce Intefeence in MANETs Hwee Xian TAN 1,2 and Winston K. G. SEAH 2,1 {stuhxt, winston}@i2.a-sta.edu.sg 1 Depatment of Compute Science, School of Computing, National Univesity

More information

(1) W tcp = (3) N. Assuming 1 P r 1. = W r (4) a 1/(k+1) W 2/(k+1)

(1) W tcp = (3) N. Assuming 1 P r 1. = W r (4) a 1/(k+1) W 2/(k+1) 1 Multi Path PERT Ankit Singh and A. L. Naasimha Reddy Electical and Compute Engineeing Depatment, Texas A&M Univesity; email: eddy@ece.tamu.edu. Abstact This pape pesents a new multipath delay based algoithm,

More information

Lifetime and Energy Hole Evolution Analysis in Data-Gathering Wireless Sensor Networks

Lifetime and Energy Hole Evolution Analysis in Data-Gathering Wireless Sensor Networks 788 IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS, VOL. 12, NO. 2, APRIL 2016 Lifetime and Enegy Hole Evolution Analysis in Data-Gatheing Wieless Senso Netwoks Ju Ren, Student Membe, IEEE, Yaoxue Zhang,

More information

IP Multicast Simulation in OPNET

IP Multicast Simulation in OPNET IP Multicast Simulation in OPNET Xin Wang, Chien-Ming Yu, Henning Schulzinne Paul A. Stipe Columbia Univesity Reutes Depatment of Compute Science 88 Pakway Dive South New Yok, New Yok Hauppuage, New Yok

More information

SCALABLE ENERGY EFFICIENT AD-HOC ON DEMAND DISTANCE VECTOR (SEE-AODV) ROUTING PROTOCOL IN WIRELESS MESH NETWORKS

SCALABLE ENERGY EFFICIENT AD-HOC ON DEMAND DISTANCE VECTOR (SEE-AODV) ROUTING PROTOCOL IN WIRELESS MESH NETWORKS SCALABL NRGY FFICINT AD-HOC ON DMAND DISTANC VCTOR (S-AODV) ROUTING PROTOCOL IN WIRLSS MSH NTWORKS Sikande Singh Reseach Schola, Depatment of Compute Science & ngineeing, Punjab ngineeing College (PC),

More information

A New Finite Word-length Optimization Method Design for LDPC Decoder

A New Finite Word-length Optimization Method Design for LDPC Decoder A New Finite Wod-length Optimization Method Design fo LDPC Decode Jinlei Chen, Yan Zhang and Xu Wang Key Laboatoy of Netwok Oiented Intelligent Computation Shenzhen Gaduate School, Habin Institute of Technology

More information

Configuring RSVP-ATM QoS Interworking

Configuring RSVP-ATM QoS Interworking Configuing RSVP-ATM QoS Intewoking Last Updated: Januay 15, 2013 This chapte descibes the tasks fo configuing the RSVP-ATM QoS Intewoking featue, which povides suppot fo Contolled Load Sevice using RSVP

More information

Improved Utility-based Congestion Control for Delay-Constrained Communication

Improved Utility-based Congestion Control for Delay-Constrained Communication Impoved Utility-based Congestion Contol fo Delay-Constained Communication Stefano D Aonco, Laua Toni, Segio Mena, Xiaoqing Zhu, and Pascal Fossad axiv:5.2799v3 [cs.ni] 2 Jan 27 Abstact Due to the pesence

More information

An Optimised Density Based Clustering Algorithm

An Optimised Density Based Clustering Algorithm Intenational Jounal of Compute Applications (0975 8887) Volume 6 No.9, Septembe 010 An Optimised Density Based Clusteing Algoithm J. Hencil Pete Depatment of Compute Science St. Xavie s College, Palayamkottai,

More information

Conservation Law of Centrifugal Force and Mechanism of Energy Transfer Caused in Turbomachinery

Conservation Law of Centrifugal Force and Mechanism of Energy Transfer Caused in Turbomachinery Poceedings of the 4th WSEAS Intenational Confeence on luid Mechanics and Aeodynamics, Elounda, Geece, August 1-3, 006 (pp337-34) Consevation Law of Centifugal oce and Mechanism of Enegy Tansfe Caused in

More information

Method of controlling access to intellectual switching nodes of telecommunication networks and systems

Method of controlling access to intellectual switching nodes of telecommunication networks and systems ISSN (e): 2250 3005 Volume 05 Issue 05 ay 2015 Intenational Jounal of Computational Engineeing eseach (IJCE) ethod of contolling access to intellectual switching nodes of telecommunication netwoks and

More information

A Recommender System for Online Personalization in the WUM Applications

A Recommender System for Online Personalization in the WUM Applications A Recommende System fo Online Pesonalization in the WUM Applications Mehdad Jalali 1, Nowati Mustapha 2, Ali Mamat 2, Md. Nasi B Sulaiman 2 Abstact foeseeing of use futue movements and intentions based

More information

Loss-Cognizant Pricing in Feed-Forward Networks with Greedy Users

Loss-Cognizant Pricing in Feed-Forward Networks with Greedy Users Loss-Cognizant Picing in Feed-Fowad Netwoks with Geedy Uses Ashaf Al Daoud and Muat Alanyali Depatment of Electical and Compute Engineeing Boston Univesity, Boston, MA 02215 Email: {ashaf, alanyali}@bu.edu

More information

Accurate Diffraction Efficiency Control for Multiplexed Volume Holographic Gratings. Xuliang Han, Gicherl Kim, and Ray T. Chen

Accurate Diffraction Efficiency Control for Multiplexed Volume Holographic Gratings. Xuliang Han, Gicherl Kim, and Ray T. Chen Accuate Diffaction Efficiency Contol fo Multiplexed Volume Hologaphic Gatings Xuliang Han, Gichel Kim, and Ray T. Chen Micoelectonic Reseach Cente Depatment of Electical and Compute Engineeing Univesity

More information

Any modern computer system will incorporate (at least) two levels of storage:

Any modern computer system will incorporate (at least) two levels of storage: 1 Any moden compute system will incopoate (at least) two levels of stoage: pimay stoage: andom access memoy (RAM) typical capacity 32MB to 1GB cost pe MB $3. typical access time 5ns to 6ns bust tansfe

More information

An effective hop-by-hop Interest shaping mechanism for CCN communications

An effective hop-by-hop Interest shaping mechanism for CCN communications An effective hop-by-hop Inteest shaping mechanism fo CCN communications Natalya Rozhnova, Sege Fdida To cite this vesion: Natalya Rozhnova, Sege Fdida. An effective hop-by-hop Inteest shaping mechanism

More information

DEADLOCK AVOIDANCE IN BATCH PROCESSES. M. Tittus K. Åkesson

DEADLOCK AVOIDANCE IN BATCH PROCESSES. M. Tittus K. Åkesson DEADLOCK AVOIDANCE IN BATCH PROCESSES M. Tittus K. Åkesson Univesity College Boås, Sweden, e-mail: Michael.Tittus@hb.se Chalmes Univesity of Technology, Gothenbug, Sweden, e-mail: ka@s2.chalmes.se Abstact:

More information

On the throughput-cost tradeoff of multi-tiered optical network architectures

On the throughput-cost tradeoff of multi-tiered optical network architectures 1 On the thoughput-cost tadeoff of multi-tieed optical netwok achitectues Guy Weichenbeg, Vincent W. S. Chan, and Muiel Médad Laboatoy fo Infomation and Decision Systems Massachusetts Institute of Technology

More information

Illumination methods for optical wear detection

Illumination methods for optical wear detection Illumination methods fo optical wea detection 1 J. Zhang, 2 P.P.L.Regtien 1 VIMEC Applied Vision Technology, Coy 43, 5653 LC Eindhoven, The Nethelands Email: jianbo.zhang@gmail.com 2 Faculty Electical

More information

Erasure-Coding Based Routing for Opportunistic Networks

Erasure-Coding Based Routing for Opportunistic Networks Easue-Coding Based Routing fo Oppotunistic Netwoks Yong Wang, Sushant Jain, Magaet Matonosi, Kevin Fall Pinceton Univesity, Univesity of Washington, Intel Reseach Bekeley ABSTRACT Routing in Delay Toleant

More information

MANET QoS support without reservations

MANET QoS support without reservations SECURITY AND COMMUNICATION NETWORKS Secuity Comm. Netwoks (2010) Published online in Wiley InteScience (www.intescience.wiley.com)..183 SPECIAL ISSUE PAPER MANET QoS suppot without esevations Soon Y. Oh,

More information

Computer Networks. TCP Libra: Derivation, analysis, and comparison with other RTT-fair TCPs

Computer Networks. TCP Libra: Derivation, analysis, and comparison with other RTT-fair TCPs Compute Netwoks 54 (2010) 2327 2344 Contents lists available at ScienceDiect Compute Netwoks jounal homepage: www.elsevie.com/locate/comnet TCP Liba: Deivation, analysis, and compaison with othe RTT-fai

More information

Scaling Location-based Services with Dynamically Composed Location Index

Scaling Location-based Services with Dynamically Composed Location Index Scaling Location-based Sevices with Dynamically Composed Location Index Bhuvan Bamba, Sangeetha Seshadi and Ling Liu Distibuted Data Intensive Systems Laboatoy (DiSL) College of Computing, Geogia Institute

More information

Exploring non-typical memcache architectures for decreased latency and distributed network usage.

Exploring non-typical memcache architectures for decreased latency and distributed network usage. Syacuse Univesity SURFACE Electical Engineeing and Compute Science Technical Repots College of Engineeing and Compute Science 9-5-20 Exploing non-typical memcache achitectues fo deceased latency and distibuted

More information

COEN-4730 Computer Architecture Lecture 2 Review of Instruction Sets and Pipelines

COEN-4730 Computer Architecture Lecture 2 Review of Instruction Sets and Pipelines 1 COEN-4730 Compute Achitectue Lectue 2 Review of nstuction Sets and Pipelines Cistinel Ababei Dept. of Electical and Compute Engineeing Maquette Univesity Cedits: Slides adapted fom pesentations of Sudeep

More information

On the Forwarding Area of Contention-Based Geographic Forwarding for Ad Hoc and Sensor Networks

On the Forwarding Area of Contention-Based Geographic Forwarding for Ad Hoc and Sensor Networks On the Fowading Aea of Contention-Based Geogaphic Fowading fo Ad Hoc and Senso Netwoks Dazhi Chen Depatment of EECS Syacuse Univesity Syacuse, NY dchen@sy.edu Jing Deng Depatment of CS Univesity of New

More information

Module 6 STILL IMAGE COMPRESSION STANDARDS

Module 6 STILL IMAGE COMPRESSION STANDARDS Module 6 STILL IMAE COMPRESSION STANDARDS Lesson 17 JPE-2000 Achitectue and Featues Instuctional Objectives At the end of this lesson, the students should be able to: 1. State the shotcomings of JPE standad.

More information

Cardiac C-Arm CT. SNR Enhancement by Combining Multiple Retrospectively Motion Corrected FDK-Like Reconstructions

Cardiac C-Arm CT. SNR Enhancement by Combining Multiple Retrospectively Motion Corrected FDK-Like Reconstructions Cadiac C-Am CT SNR Enhancement by Combining Multiple Retospectively Motion Coected FDK-Like Reconstuctions M. Pümme 1, L. Wigstöm 2,3, R. Fahig 2, G. Lauitsch 4, J. Honegge 1 1 Institute of Patten Recognition,

More information

INFORMATION DISSEMINATION DELAY IN VEHICLE-TO-VEHICLE COMMUNICATION NETWORKS IN A TRAFFIC STREAM

INFORMATION DISSEMINATION DELAY IN VEHICLE-TO-VEHICLE COMMUNICATION NETWORKS IN A TRAFFIC STREAM INFORMATION DISSEMINATION DELAY IN VEHICLE-TO-VEHICLE COMMUNICATION NETWORKS IN A TRAFFIC STREAM LiLi Du Depatment of Civil, Achitectual, and Envionmental Engineeing Illinois Institute of Technology 3300

More information

Minimizing Application-Level Delay of Multi-Path TCP in Wireless networks: A Receiver-Centric Approach

Minimizing Application-Level Delay of Multi-Path TCP in Wireless networks: A Receiver-Centric Approach Minimizing Application-Level Delay of Multi-Path TCP in Wieless netwoks: A Receive-Centic Appoach Se-Yong Pak, Changhee Joo, Yongseok Pak, and Saewoong Bahk Depament of ECE and INMC, Seoul National Univesity,

More information

An Extension to the Local Binary Patterns for Image Retrieval

An Extension to the Local Binary Patterns for Image Retrieval , pp.81-85 http://x.oi.og/10.14257/astl.2014.45.16 An Extension to the Local Binay Pattens fo Image Retieval Zhize Wu, Yu Xia, Shouhong Wan School of Compute Science an Technology, Univesity of Science

More information

A Novel Automatic White Balance Method For Digital Still Cameras

A Novel Automatic White Balance Method For Digital Still Cameras A Novel Automatic White Balance Method Fo Digital Still Cameas Ching-Chih Weng 1, Home Chen 1,2, and Chiou-Shann Fuh 3 Depatment of Electical Engineeing, 2 3 Gaduate Institute of Communication Engineeing

More information

TCP Libra: Exploring RTT-Fairness for TCP

TCP Libra: Exploring RTT-Fairness for TCP TCP Liba: Exploing RTT-Fainess fo TCP UCLA Compute Science Depatment Technical Repot #TR050037 Gustavo Mafia, Claudio Palazzi, Giovanni Pau, Maio Gela, M. Y. Sanadidi, Maco Roccetti, Compute Science Depatment

More information

Color Correction Using 3D Multiview Geometry

Color Correction Using 3D Multiview Geometry Colo Coection Using 3D Multiview Geomety Dong-Won Shin and Yo-Sung Ho Gwangju Institute of Science and Technology (GIST) 13 Cheomdan-gwagio, Buk-ku, Gwangju 500-71, Republic of Koea ABSTRACT Recently,

More information

DYNAMIC STORAGE ALLOCATION. Hanan Samet

DYNAMIC STORAGE ALLOCATION. Hanan Samet ds0 DYNAMIC STORAGE ALLOCATION Hanan Samet Compute Science Depatment and Cente fo Automation Reseach and Institute fo Advanced Compute Studies Univesity of Mayland College Pak, Mayland 07 e-mail: hjs@umiacs.umd.edu

More information

A modal estimation based multitype sensor placement method

A modal estimation based multitype sensor placement method A modal estimation based multitype senso placement method *Xue-Yang Pei 1), Ting-Hua Yi 2) and Hong-Nan Li 3) 1),)2),3) School of Civil Engineeing, Dalian Univesity of Technology, Dalian 116023, China;

More information

Topological Characteristic of Wireless Network

Topological Characteristic of Wireless Network Topological Chaacteistic of Wieless Netwok Its Application to Node Placement Algoithm Husnu Sane Naman 1 Outline Backgound Motivation Papes and Contibutions Fist Pape Second Pape Thid Pape Futue Woks Refeences

More information

Frequency Domain Approach for Face Recognition Using Optical Vanderlugt Filters

Frequency Domain Approach for Face Recognition Using Optical Vanderlugt Filters Optics and Photonics Jounal, 016, 6, 94-100 Published Online August 016 in SciRes. http://www.scip.og/jounal/opj http://dx.doi.og/10.436/opj.016.68b016 Fequency Domain Appoach fo Face Recognition Using

More information

ADDING REALISM TO SOURCE CHARACTERIZATION USING A GENETIC ALGORITHM

ADDING REALISM TO SOURCE CHARACTERIZATION USING A GENETIC ALGORITHM ADDING REALISM TO SOURCE CHARACTERIZATION USING A GENETIC ALGORITHM Luna M. Rodiguez*, Sue Ellen Haupt, and Geoge S. Young Depatment of Meteoology and Applied Reseach Laboatoy The Pennsylvania State Univesity,

More information

Minimizing spatial and time reservation with Collision-Aware DCF in mobile ad hoc networks

Minimizing spatial and time reservation with Collision-Aware DCF in mobile ad hoc networks Available online at www.sciencediect.com Ad Hoc Netwoks 7 (29) 23 247 www.elsevie.com/locate/adhoc Minimizing spatial and time esevation with Collision-Awae DCF in mobile ad hoc netwoks Lubo Song a, Chansu

More information

Fault-Tolerant Routing Schemes in RDT(2,2,1)/α-Based Interconnection Network for Networks-on-Chip Designs

Fault-Tolerant Routing Schemes in RDT(2,2,1)/α-Based Interconnection Network for Networks-on-Chip Designs Fault-Toleant Routing Schemes in RDT(,,)/α-Based Inteconnection Netwok fo Netwoks-on-Chip Designs Mei Yang, Tao Li, Yingtao Jiang, and Yulu Yang Dept. of Electical & Compute Engineeing Univesity of Nevada,

More information

Dynamic Processor Scheduling with Client Resources for Fast Multi-resolution WWW Image Browsing

Dynamic Processor Scheduling with Client Resources for Fast Multi-resolution WWW Image Browsing Dynamic Pocesso Scheduling with Resouces fo Fast Multi-esolution WWW Image Bowsing Daniel Andesen, Tao Yang, David Watson, and Athanassios Poulakidas Depatment of Compute Science Univesity of Califonia

More information

COSC 6385 Computer Architecture. - Pipelining

COSC 6385 Computer Architecture. - Pipelining COSC 6385 Compute Achitectue - Pipelining Sping 2012 Some of the slides ae based on a lectue by David Culle, Pipelining Pipelining is an implementation technique wheeby multiple instuctions ae ovelapped

More information

An Efficient Handover Mechanism Using the General Switch Management Protocol on a Multi-Protocol Label Switching Network

An Efficient Handover Mechanism Using the General Switch Management Protocol on a Multi-Protocol Label Switching Network An Efficient andove Mechanism Using the Geneal Switch Management Potocol on a Multi-Potocol abel Switching Netwok Seong Gon hoi, yun Joo Kang, and Jun Kyun hoi Using the geneal switch management potocol

More information

Title. Author(s)NOMURA, K.; MOROOKA, S. Issue Date Doc URL. Type. Note. File Information

Title. Author(s)NOMURA, K.; MOROOKA, S. Issue Date Doc URL. Type. Note. File Information Title CALCULATION FORMULA FOR A MAXIMUM BENDING MOMENT AND THE TRIANGULAR SLAB WITH CONSIDERING EFFECT OF SUPPO UNIFORM LOAD Autho(s)NOMURA, K.; MOROOKA, S. Issue Date 2013-09-11 Doc URL http://hdl.handle.net/2115/54220

More information

Approximating Euclidean Distance Transform with Simple Operations in Cellular Processor Arrays

Approximating Euclidean Distance Transform with Simple Operations in Cellular Processor Arrays 00 th Intenational Wokshop on Cellula Nanoscale Netwoks and thei Applications (CNNA) Appoximating Euclidean Distance Tansfom with Simple Opeations in Cellula Pocesso Aas Samad Razmjooei and Piot Dudek

More information

= dv 3V (r + a 1) 3 r 3 f(r) = 1. = ( (r + r 2

= dv 3V (r + a 1) 3 r 3 f(r) = 1. = ( (r + r 2 Random Waypoint Model in n-dimensional Space Esa Hyytiä and Joma Vitamo Netwoking Laboatoy, Helsinki Univesity of Technology, Finland Abstact The andom waypoint model (RWP) is one of the most widely used

More information

Gravitational Shift for Beginners

Gravitational Shift for Beginners Gavitational Shift fo Beginnes This pape, which I wote in 26, fomulates the equations fo gavitational shifts fom the elativistic famewok of special elativity. Fist I deive the fomulas fo the gavitational

More information

High performance CUDA based CNN image processor

High performance CUDA based CNN image processor High pefomance UDA based NN image pocesso GEORGE VALENTIN STOIA, RADU DOGARU, ELENA RISTINA STOIA Depatment of Applied Electonics and Infomation Engineeing Univesity Politehnica of Buchaest -3, Iuliu Maniu

More information

Matching Schemes with Captured-Frame Eligibility for Input-Queued Packet Switches

Matching Schemes with Captured-Frame Eligibility for Input-Queued Packet Switches Matching Schemes with Captured-Frame Eligibility for -Queued Packet Switches Roberto Rojas-Cessa and Chuan-bi Lin Abstract Virtual output queues (VOQs) are widely used by input-queued (IQ) switches to

More information

Reachable State Spaces of Distributed Deadlock Avoidance Protocols

Reachable State Spaces of Distributed Deadlock Avoidance Protocols Reachable State Spaces of Distibuted Deadlock Avoidance Potocols CÉSAR SÁNCHEZ and HENNY B. SIPMA Stanfod Univesity We pesent a family of efficient distibuted deadlock avoidance algoithms with applications

More information

Lecture 8 Introduction to Pipelines Adapated from slides by David Patterson

Lecture 8 Introduction to Pipelines Adapated from slides by David Patterson Lectue 8 Intoduction to Pipelines Adapated fom slides by David Patteson http://www-inst.eecs.bekeley.edu/~cs61c/ * 1 Review (1/3) Datapath is the hadwae that pefoms opeations necessay to execute pogams.

More information

Journal of Network and Computer Applications

Journal of Network and Computer Applications Jounal of Netwok and Compute Applications 34 (211) 135 142 Contents lists available at ScienceDiect Jounal of Netwok and Compute Applications jounal homepage: www.elsevie.com/locate/jnca Optimization of

More information

On the Conversion between Binary Code and Binary-Reflected Gray Code on Boolean Cubes

On the Conversion between Binary Code and Binary-Reflected Gray Code on Boolean Cubes On the Convesion between Binay Code and BinayReflected Gay Code on Boolean Cubes The Havad community has made this aticle openly available. Please shae how this access benefits you. You stoy mattes Citation

More information

The concept of PARPS - Packet And Resource Plan Scheduling

The concept of PARPS - Packet And Resource Plan Scheduling The concept of PARPS - Packet And Resouce Plan Scheduling Magnus Eiksson 1 and Håkan Sätebeg 2 1) Dept. of Signals, Sensos and Systems, Royal Inst. of Technology, Sweden. E-mail: magnus.eiksson@ite.mh.se.

More information

Clustering Interval-valued Data Using an Overlapped Interval Divergence

Clustering Interval-valued Data Using an Overlapped Interval Divergence Poc. of the 8th Austalasian Data Mining Confeence (AusDM'9) Clusteing Inteval-valued Data Using an Ovelapped Inteval Divegence Yongli Ren Yu-Hsn Liu Jia Rong Robet Dew School of Infomation Engineeing,

More information

A Fuzzy Constraint-Based Routing Algorithm for Traffic Engineering

A Fuzzy Constraint-Based Routing Algorithm for Traffic Engineering Fuzz Constaint-ased Routing lgoithm fo Taffic Engineeing Junaid. Khan and Hussein M. lnuweii Depatment of Electical & Compute Engineeing, 2356 Main Mall, Univesit of itish Columbia, Vancouve,.C. Canada,

More information

Segmentation of Casting Defects in X-Ray Images Based on Fractal Dimension

Segmentation of Casting Defects in X-Ray Images Based on Fractal Dimension 17th Wold Confeence on Nondestuctive Testing, 25-28 Oct 2008, Shanghai, China Segmentation of Casting Defects in X-Ray Images Based on Factal Dimension Jue WANG 1, Xiaoqin HOU 2, Yufang CAI 3 ICT Reseach

More information

A Shape-preserving Affine Takagi-Sugeno Model Based on a Piecewise Constant Nonuniform Fuzzification Transform

A Shape-preserving Affine Takagi-Sugeno Model Based on a Piecewise Constant Nonuniform Fuzzification Transform A Shape-peseving Affine Takagi-Sugeno Model Based on a Piecewise Constant Nonunifom Fuzzification Tansfom Felipe Fenández, Julio Gutiéez, Juan Calos Cespo and Gacián Tiviño Dep. Tecnología Fotónica, Facultad

More information

A Full-mode FME VLSI Architecture Based on 8x8/4x4 Adaptive Hadamard Transform For QFHD H.264/AVC Encoder

A Full-mode FME VLSI Architecture Based on 8x8/4x4 Adaptive Hadamard Transform For QFHD H.264/AVC Encoder 20 IEEE/IFIP 9th Intenational Confeence on VLSI and System-on-Chip A Full-mode FME VLSI Achitectue Based on 8x8/ Adaptive Hadamad Tansfom Fo QFHD H264/AVC Encode Jialiang Liu, Xinhua Chen College of Infomation

More information

Achievable Catalog Size in Peer-to-Peer Video-on-Demand Systems

Achievable Catalog Size in Peer-to-Peer Video-on-Demand Systems Achievable Catalog Size in Pee-to-Pee Video-on-Demand Systems Yacine Boufkhad, Fabien Mathieu, Fabien De Montgolfie, Diego Peino, Lauent Viennot To cite this vesion: Yacine Boufkhad, Fabien Mathieu, Fabien

More information