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

Size: px
Start display at page:

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

Transcription

1 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 Abstact Numeous solutions have been poposed in the liteatue to eliminate eodeing in load-balanced switch fabics. A common appoach involves unifom fames, in which evey cell of a fame has the same destination. This can achieve 100% thoughput with elatively small aveage taffic delay; howeve, the wost-case delay may be unbounded. We show that with a slight speedup in the switch fabic we can guaantee satisfactoy wost-case delay bounds, without sacificing othe desiable popeties. Futhemoe, expeimental esults demonstate that ou scheme impoves wost-case delay in ealistic taffic scenaios, as compaed to pevious unifom-fame solutions. I. INTRODUCTION Load balancing in switch fabics has been studied extensively fo moe than a decade, as a building block to obtain high-capacity outes; see e.g. [1], [2], [3], [7], [8], [9], [12], among othes. The pimay motivation behind these eseach effots is the explosive gowth in Intenet taffic, vaiously estimated at 30-50% pe yea [4], [10]. As a esult of that tend, load-balanced switch fabics have become inceasingly impotant because they ae inheently paallel, enabling switch capacity to scale even without any incease in VLSI clock ate. Futhemoe, load-balanced switch fabics ae known to attain 100% thoughput fo a easonably geneal model of admissible taffic [1], [2]. A typical load-balanced switch fabic consists of thee stages of cell-handling elements. In this pape, we focus on fame-based fabics with a full-mesh inteconnect between successive stages, as illustated in Figue 1. Data packets, aleady split into fixed-size cells, aive at a fist-stage distibuto. Cells ae accumulated into a fame of cells that ae sent out in paallel to the middle stage (called the outing stage), one cell pe outing element. Each outing element similaly accumulates fames of cells to be sent out in paallel to the final collecto stage. Nevetheless, unlike a distibuto, each outing stage pefoms outing, so that each cell is sent to its intended collecto. Fom thee cells exit the switch fabic and ae then eassembled into packets. Fame-based fabics allow high-aggegate-capacity switches to be built fom elements that need not opeate much faste than the ate at which taffic can aive at a distibuto. Moeove, if thee ae m elements at each stage, each link between stages opeates at a faction 1/m of the ate into the distibuto. Ideally, the m 2 links between stages could be encapsulated in a single optical device, a polycyclic aay wave guide [9]. We emak that load-balanced switch fabics wee initially descibed with cossba-based inteconnects ealizing distibutos Fig. 1. mesh mesh outing elements Fame-based switch fabic collectos a fixed schedule of matchings [1], [2]; the full mesh descibed above eplaces the cossba, which would be challenging to implement at high ates. Cells fom a specific distibuto sent to a specific collecto may tavese diffeent outing elements. If these outing elements do not impose the same delay (e.g. because of diffeing queue occupancies), cells ae likely to aive at the collecto in a diffeent ode fom which they wee sent. Requiing that fames be unifom, i.e. all cells in each fame have the same destination, has been poposed as a way to pevent this celleodeing poblem [7], [8]. Indeed, unifom fames guaantee that, at any given time and fo any given collecto, all outing element queues to that collecto have identical occupancies and thus incu identical delays. In this way, cell eodeing absolutely cannot occu, since each fame aives at its destined collecto as a whole, exactly as it was sent fom the distibuto, and of couse no fame can ovetake an ealie fame sent to the same collecto. Seveal switch scheduling policies that ae based on the afoementioned appoach [7], [8] povably achieve 100% thoughput and ensue bounded aveage cell delay. Unfotunately, though, thee is no guaantee whatsoeve in tems of the wost-case delay, which can be abitaily lage. Poviding such a guaantee is the main goal of this pape. Moe specifically, we pesent the Bounded Latency Fames (BLF) scheduling policy, which pemits a fame to be sent fom a distibuto if: (a) thee ae enough cells with the same destination to fill a fame; o (b) some cell in the distibuto has been waiting fo moe than a specific peiod of time. We subsequently show that, fo any taffic aival patten, a unifom-fame-based switch fabic with this scheduling policy

2 2 delays a cell by at most a constant moe than would an ideal output-queued switch. To obtain the above esult, we equie that the switch fabic have a slight speedup, because sending fames befoe they ae completely filled wastes bandwidth. Additionally, we assume that intenal switch queues do not have maximum occupancy limits and may gow abitaily lage which is standad fo such theoetical compaisons. Theefoe, ou esult complements those of [7], [8]: a stonge assumption (namely, fabic speedup) yields a stonge guaantee (i.e., wost-case bounded delay athe than aveage-case bounded delay). The pape is oganized as follows. Section II descibes ou BLF scheduling policy and fomally states the pefomance guaantee that BLF povides. All technical poofs, howeve, ae elegated to Section III. In Section IV we evaluate the pefomance of BLF in ealistic settings via simulations. Last but not least, Section V pesents ou concluding emaks. A. Peliminaies II. BOUNDED LATENCY FRAMES As mentioned ealie, in a unifom load-balanced switch, evey fame sent by a distibuto is unifom, that is, evey cell in the fame has the same collecto as destination. Conside the schematic epesentation of Figue 1, and let m denote the numbe of elements on each of the thee stages (distibutos, outing elements, and collectos). Then, each fame consists of exactly m cells. At fixed intevals, each of which is called a fame time, evey distibuto sends a fame to the outing stage. Fo that pupose, evey distibuto must apply the scheduling policy independently to detemine the collecto fo which each fame is intended. If necessay, empty cells containing dummy data may be added to a fame to each the equisite numbe of m cells so that it can be sent. Similaly, at each fame time, evey outing element sends at most one cell to each collecto. Routing elements opeate deteministically, all using the same algoithm to fill queues. Fo example, if at a given fame time multiple distibutos send fames destined to the same collecto, then the ode in which the outing elements queues to that collecto ae filled with cells fom those fames is identical fo all elements (e.g. beak ties by the index of each fame s distibuto). Consequently, at any time, all the outing-stage queues associated with a specific collecto have the same length, and futhemoe all cells associated with a paticula distibuto fame ae at the same position in these queues. Thus, all cells in a fame aive at thei destination simultaneously. Note that an empty cell used to fill out a fame occupies a position within a outing element queue, and pesists until it eaches a collecto, whee it is dopped. As an exception to the above, a fame consisting only of empty cells 1 should be maked by the distibuto so that it can be discaded immediately afte aiving at the outing stage. 1 Sending such fames seems pointless, but in pactice it could be equied to maintain fabic synchonization. Finally, we assume that duing each fame time at most m = α cells may aive at a distibuto, and also at most cells may depat a collecto. The quantity α > 1 expesses the speedup of the switch fabic compaed to the capacity of the input and output links at the distibutos and collectos, espectively. B. Scheduling policy A vey obvious pefomance concen in unifom switch fabics is the delay incued while a distibuto waits fo enough cells with the same destination to ceate a fame. Filling up fames with empty cells enables them to be sent in a moe timely fashion, but doing this too often consumes a lot of bandwidth and leads to high queueing delays in the outing stage. Nevetheless, by applying the above technique judiciously and exploiting the speedup α, we demonstate a scheduling policy that bounds wost-case delay. As cells aive at a distibuto, they ae placed in the appopiate vitual output queue by destination, with the oldest cell at the head of the queue. At any given time, the cells in a queue ae patitioned in ode fom oldest to youngest into (potential) fames: zeo o moe full fames containing exactly m cells, and possibly a patial fame containing at least one but fewe than m cells. Now, define = αm α 1. (1) A patial fame is stale if the cuent fame time exceeds the aival time of the oldest cell in the fame by at least. In the Bounded Latency Fames (BLF) scheduling policy, each distibuto examines its vitual output queues in oundobin ode, choosing a queue to seve if it contains eithe a full fame o a stale fame. Thus, the fame at the head of the chosen queue is sent to the outing stage, padded with empty cells if it is stale. If no queue has a full o stale fame, a fame of all empty cells is sent instead. Lemma 1. Unde the BLF scheduling policy, no cell is delayed at a distibuto fo moe than + m fame times. We have aleady pointed out that sending stale fames squandes bandwidth. To establish that the speedup α suffices to make up fo that loss, we elate the delay expeienced by an abitay cell though the unifom switch fabic with that of an ideal output-queued switch. In the latte, at each fame time, any cells that aive at the distibutos ae instantaneously fowaded to the appopiate collecto, and then cells depat fom each collecto queue o the entie queue, if its occupancy is less than. An aival sequence {a t } (fo a specific collecto) indicates the total numbe a t of cells aiving ove all distibutos destined fo the collecto duing the t-th fame inteval. The intinsic backlog b t (fo a specific collecto) at time t fo an aival sequence {a t } is defined as ( ) f b t = max 0, max (a t i ). 0 f t i=0

3 3 It is easy to see that b t expesses the numbe of cells in the collecto queue of an ideal output-queued switch at time t, afte the cells have been emitted. This can also be veified by an inductive agument: it is obviously tue at time 0, and at time t the numbe of cells in the queue is max(0, b t 1 +a t ), which equals b t. Moeove, we define the intinsic delay δ t at time t as δ t = bt Clealy, δ t is the delay expeienced by the last cell to aive at the output-queued switch duing fame inteval t. Theoem 2. Suppose a unifom switch fabic with the BLF scheduling policy has unbounded intenal queues. Then, a cell aiving at a distibuto at time t will exit the collecto within δ t+ +m m fame intevals. In othe wods, fo any aival sequence, the delay seen by a cell tavesing the unifom switch fabic is only up to a constant (depending on m and α) geate than the delay expeienced by a cell aiving at almost the same time at an ideal output-queued switch. Remak. In contast to the above, suppose the switch has a congestion contol mechanism that pevents intenal queues fom gowing too lage by tempoaily focing upsteam elements to stop sending moe cells. In such a scenaio, the delay guaantees of Lemma 1 and Theoem 2 ae no longe valid even fo taffic to a collecto that is neve congested. This complication is not specific to the unifom load-balanced switch, and additional speedup is equied to addess it, even with coss-ba-based fabics [5]. C. Extensions Conside a situation in which the incoming taffic load to the switch is low, i.e. on aveage much fewe than cells aive at each distibuto pe fame time. Unde the BLF scheduling policy, these cells will expeience a delay of oughly fame times, in expectation. This leaves oom fo impovement, since the bandwidth of the switch is undeutilized. In paticula, we can educe aveage cell delay in the afoementioned setting by combining BLF with a technique inspied fom [7]. Namely, if a distibuto has no vitual output queue with a full o stale fame to seve, it again examines the queues in ound-obin ode, choosing a queue Q to seve if: (a) Q contains a patial fame; and (b) the outing-stage queues associated with the same collecto as Q cuently hold less than T fames. Hee, T is a paamete that we set as desied. Natually, if thee is still no suitable queue to seve, a fame of all empty cells is sent. As a esult, this modified BLF policy tends to decease the time that cells have to wait at a distibuto, povided that the outing-stage queues they will be sent to have sufficiently low occupancies. Moe impotantly, the wost-case delay guaantee of Theoem 2 is only minimally affected by that change. Theoem 3. Suppose a unifom switch fabic with the (modified) BLF scheduling policy has unbounded intenal queues.. Then, a cell aiving at a distibuto at time t will exit the collecto within δ t+ +m m + T fame intevals. III. PROOFS Poof of Lemma 1: We say that a fame in a distibuto queue is blocked if thee is a fame ahead of it in the queue, necessaily a full fame. Now, obseve that the lemma is staightfowad fo any fame that is not blocked, since at wost it will become stale in fame times, and will then be seved within anothe m fame times. So, let us assume that some fame is blocked at fame time t+1 (by the aival of the (m+1)-st cell to a queue), but thee wee no blocked fames at time t. We claim that at some fame time between t + 1 and t +, inclusive, thee will be no full fames, hence no blocked fames. Evey once-blocked fame will eithe have been seved o become head of its queue; in the latte case, the fame will be seved within m time units of becoming full o stale. To see the claim, fist note that thee must exist a full fame at evey fame time that thee is a blocked fame, which implies that some fame is seved then. In the fame times between t + 1 and t + δ, δ, thee ae clealy δ sevice oppotunities; how many fames will need to be seved? We ague at most m + δ/α. Indeed, thee ae at most m fames that ae patial o full at fame time t, one pe queue, but no blocked fames. Moeove, if a patial fame is seved because it became stale, no othe stale fame fom the same queue is seved duing that inteval. As a esult, any othe fame that is seved must have all its cells aive between fame times t and t + δ. Thee can be at most δ = δ m/α such cells, hence at most δ/α full fames. Now, since m + δ/α (α 1) /α + /α, thee exists some δ such that, by fame time t + δ, thee will have been as many sevice oppotunities as fames to seve, and all full fames will have been seved. Fo the following poofs, we only count queueing delays and ignoe any delays due to pocessing o tansfes between stages. Thus, a cell that aives at a distibuto at fame time t could be emitted fom a collecto also at time t if it wee not delayed because of queue occupancy. A collecto queue is backlogged at time t if it is not empty afte the depatue of the cells at time t. Suppose a cell is emitted fom a collecto queue at time t. The collecto backlog inteval of the cell is the inteval [s..t 1] such that the queue is backlogged at times s,..., t 1, but not backlogged at time s 1. Likewise, the outing-stage queues fo a specific collecto ae backlogged at time t if they contain anothe fame to be sent to the collecto (containing at least one nonempty cell). Futhemoe, the outing backlog inteval of a cell is defined similaly to the collecto backlog inteval. Poposition 4. Suppose the outing-stage queues fo a specific collecto ae backlogged fo an inteval I of g consecutive fame times, but not backlogged befoe I. Then, the fames sent

4 4 E D t C t B A distibuto outing stage collecto Fig. 2. Time lines (hoizontal) in ode fom top to bottom: aival time at distibuto; depatue fom distibuto/aival at outing stage; depatue fom outing stage/aival at collecto; depatue fom collecto. The diagonal dotted line epesents the tajectoy of cell c aiving at time t. to the collecto duing I contain at least (g m) nonempty cells. Poof: Since the outing stages wee not backlogged befoe I, any fames sent to the collecto must have aived duing I. Wite g = g 1 + g 2, whee g 1 = g/ and g 2 = g mod. At most g stale fames can aive at the outing stage duing the inteval fom a single distibuto. Howeve, thee can be at most min(g 2, m) such distibutos, since the aival times at the outing stage of the fist and last stale fames ae sepaated by at least g 1. Othe distibutos must contibute g 1 o fewe stale fames each. Hence, the total numbe of stale fames summed ove all distibutos is g 1 m + min(g 2, m). To finish the poof, the numbe η of nonempty cells is at least m times the numbe of full fames: η m (g ( g 1 m + min(g 2, m) )) ( mα α g 1 α 1 + g 2 g ) 1m(α 1) min(g 2, m) α 1 ( ) g1 mα = α 1 + g 2 + (α 1)g 2 α min(g 2, m) (g m), whee the inequality (α 1)g 2 α min(g 2, m) m is easily veified using a case split on g 2 m o g 2 > m. Poof of Theoem 2: Conside cell c aiving at a distibuto at time t, destined fo a specific collecto. Define intevals of fame times A, B, C, D, and E as follows. Let A = [s A..t A 1] be the collecto backlog inteval of c. B A is the inteval fom the stat of A, i.e. s A, to the aival time of c at the collecto. Obviously, at time s A one o moe cells must aive at this collecto fom the outing stage, othewise the collecto would not become backlogged. The outing backlog inteval of these cells is C, which may be empty. D B C is the inteval fom the stat of C (o of B, if C empty) to the aival of c at the outing stage. Finally, E is the inteval of length + m befoe D. Figue 2 povides a visualization of the above definitions. We wish to bound the numbe a of cells that ae emitted fom the collecto duing A. Any such cell must have aived at the collecto duing B, since A is the backlog inteval of c. Likewise, any cell that aives at the collecto duing B C must have aived at the outing stage duing D. By Poposition 4, the numbe of cells sent to the collecto duing C is at least ( C m), whee C is the length of C. By Lemma 1, any cell sent to the outing stage duing D must have aived at the distibuto duing D E. Consequently, a a s ( C m). We have A = a/ since A is the backlog inteval fo c. Let t be the endpoint of D. Using E = + m, 1 A + C D a s + m D 1 = (a s ) + + 2m The cell is emitted at time δ t + + 2m. t A = t + (t t) + A + C D t + (t t + δ t ) + + 2m t + ( + m + δ t+ +m ) + + 2m, whee the second inequality uses the obsevation that the quantity t t + δ t cannot decease as t vaies fom t to t + + m. Sketch of poof of Theoem 3: The poof is vey simila to that of Theoem 2, so hee we only highlight thei diffeences. To begin with, obseve that Lemma 1 holds fo the modified BLF policy as well. Moeove, let us amend two definitions fom Section III. In paticula, the outing-stage queues fo a specific collecto ae backlogged at time t if they contain moe than T fames, and the outing backlog inteval of a cell is analogously edefined. Unde the new definitions, Poposition 4 also caies ove, with a mino change in its statement: the fames sent to the collecto duing I + T (i.e. the inteval I shifted fowad by T fame times) contain at least (g m) nonempty cells. Lastly, fo the aguments in the poof of Theoem 2 to go though, we modify the intevals B and C: B now stats at time s A T instead of s A, and C is the backlog inteval of the cells, if any, that at time s A T occupy the T -th position in the outing-stage queues fo the collecto in question (which means that they will be sent to the collecto at time s A ). It is easy to deduce that the net esult of all those changes is simply an additional tem T in the delay bound. IV. SIMULATIONS To complement the theoetical analysis, we conducted pefomance simulations of the BLF scheduling policy, specifically its vesion descibed in Section II-C, unde ealistic taffic pattens. The benchmak used fo these simulations is the closely elated Padded Fames (PF) policy [7], which has

5 5 Max delay (fame times) % 20% 30% 40% 50% 60% 70% 80% 90% Taffic load PF BLF Fig. 3. Plot of maximum packet delay vesus taffic load fo the PF and BLF scheduling policies, on a switch. The hoizontal dotted line shows the value of, included fo efeence puposes. been expeimentally shown to exhibit lowe aveage delay than othe poposed unifom-fame policies. The load-balanced switch fabic on which PF and BLF wee applied had m = 32 elements pe stage, and a small speedup α = Thus, (1) yields = 147. Additionally, the paamete T was set to 4 fo both policies. Note that PF does not equie any speedup, unlike BLF; nevetheless, fo the pupose of a fai compaison we chose to keep the speedup in both cases. Futhemoe, the incoming taffic at distibuto i destined fo collecto j follows a Paeto distibution with aveage ate ρ ij. These ates ae andomly chosen fo each distibuto-collecto pai, subject to the condition ρ ij = ρ ij = l, i j fo all i and j. The paamete l above, which is independent of i and j, epesents the aveage taffic load on the switch, and in ou test instances it anges fom 10% to 90% of the capacity of each of the input and output links that cay taffic into the distibutos and out of the collectos, espectively. Figue 3 pesents the maximum obseved delay that was incued unde each scheduling policy, expessed in fame times. As expected, BLF ensues that all taffic taveses the switch in a timely manne. By contast, fo modeate to high loads, PF causes some packets to expeience vey lage delays. It is also instuctive to examine these esults in conjunction with the coesponding aveage delays, which ae displayed in Table I. Note that both policies attain almost identical aveage delays in each case, with BLF gaining a vey slight advantage as the load inceases. Consequently, it appeas that BLF s focus on bounding maximum delay (by expediently sending stale fames, which consume moe bandwidth) does not ham oveall pefomance in tems of aveage delay on the contay, it maginally impoves it. TABLE I AVERAGE PACKET DELAY FOR PF AND BLF Taffic load PF BLF 10% % % % % % % % % delay expeienced by taffic passing though the switch is at most a constant moe than the coesponding delay though an ideal output-queued switch; this distinguishes BLF fom peviously known unifom-fame policies. Moeove, in ode to be implemented, BLF equies only a small speedup in the switch fabic, e.g. in the ode of Last but not least, expeimental evidence indicates that in pactice the maximum delay bound that BLF achieves does not come at the expense of wose aveage delay, whence we conclude that BLF delives a well-ounded pefomance. REFERENCES [1] C. Chang, D. Lee, Y. Jou, Load-balanced Bikhoff-von Neumann switches: Pat I: One-stage Buffeing, Comput. Commun. 25:6, 2002, pp [2] C. Chang, D. Lee, Y. Jou, Load-balanced Bikhoff-von Neumann switches: Pat II: Multi-stage Buffeing, Comput. Commun. 25:6, 2002, pp [3] C. Chang, D. Lee, Y. Shih, C. Yu, Mailbox Switch: A Scalable two-stage Switch Achitectue fo Conflict Resolution of Odeed Packets, IEEE Tansactions on Communications, 56:1, Januay 2008, pp [4] CISCO Visual Netwoking Index: Foecase and Methodology, , available within [5] S.-T. Chuang, A. Goel, N. McKeown, B. Pabhaka, Matching Output Queueing with a Combined Input Output Queued Switch, Compute Systems Technical Repot CSL-TR Mach 1998, o Poceedings of INFOCOM 99, , IEEE, Apil 1999, o IEEE Jounal on Selected Aeas in Communications, 17:6, Decembe 1999, pp [6] W. Felle, An Intoduction to Pobability Theoy and its Applications, Volume I, Thid Edition, John Wiley and Sons, NY, [7] J. Jaamillo, F. Milan, R. Sikant, Padded Fames: a Novel Algoithm fo Stable Scheduling in Load-Balanced Switches, IEEE/ACM Tansactions on Netwoking, 16:5, Octobe 2008, pp [8] I. Keslassy, The Load-balanced Route, Ph.D. Dissetation, Stanfod Univ, Stanfod CA, [9] I. Keslassy, S.-T. Chuang, K. Yu, D. Mille, M. Hoowitz, O. Solgaad, N. McKeown, Scaling Intenet Routes Using Optics (Extended Vesion), Stanfod Technical Repot TR03-HPNG , [10] Minnesota Intenet Taffic Studies (MINTS), available at [11] M. Mitzenmache, E. Upfal, Pobability and Computing: Randomized algoithms and pobabilistic analysis, Cambidge Univesity Pess, Cambidge, [12] Y. Shen, S. Panwa, H. Chao, Design and Pefomance Analysis of a Pactical Load-Balanced Switch, IEEE Tansactions on Communications 57:8, August 2009, pp V. CONCLUSION In this pape, we intoduced a novel scheduling policy fo load-balanced switch fabics, called Bounded Latency Fames. We also poved that, unde the BLF policy, the wost-case

The Dual Round Robin Matching Switch with Exhaustive Service

The Dual Round Robin Matching Switch with Exhaustive Service 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

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

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

FACE VECTORS OF FLAG COMPLEXES

FACE VECTORS OF FLAG COMPLEXES FACE VECTORS OF FLAG COMPLEXES ANDY FROHMADER Abstact. A conjectue of Kalai and Eckhoff that the face vecto of an abitay flag complex is also the face vecto of some paticula balanced complex is veified.

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

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

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

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

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

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

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

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

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

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

RANDOM IRREGULAR BLOCK-HIERARCHICAL NETWORKS: ALGORITHMS FOR COMPUTATION OF MAIN PROPERTIES

RANDOM IRREGULAR BLOCK-HIERARCHICAL NETWORKS: ALGORITHMS FOR COMPUTATION OF MAIN PROPERTIES RANDOM IRREGULAR BLOCK-HIERARCHICAL NETWORKS: ALGORITHMS FOR COMPUTATION OF MAIN PROPERTIES Svetlana Avetisyan Mikayel Samvelyan* Matun Kaapetyan Yeevan State Univesity Abstact In this pape, the class

More information

Shortest Paths for a Two-Robot Rendez-Vous

Shortest Paths for a Two-Robot Rendez-Vous Shotest Paths fo a Two-Robot Rendez-Vous Eik L Wyntes Joseph S B Mitchell y Abstact In this pape, we conside an optimal motion planning poblem fo a pai of point obots in a plana envionment with polygonal

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

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

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 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

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

ART GALLERIES WITH INTERIOR WALLS. March 1998

ART GALLERIES WITH INTERIOR WALLS. March 1998 ART GALLERIES WITH INTERIOR WALLS Andé Kündgen Mach 1998 Abstact. Conside an at galley fomed by a polygon on n vetices with m pais of vetices joined by inteio diagonals, the inteio walls. Each inteio wall

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

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

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

Point-Biserial Correlation Analysis of Fuzzy Attributes

Point-Biserial Correlation Analysis of Fuzzy Attributes Appl Math Inf Sci 6 No S pp 439S-444S (0 Applied Mathematics & Infomation Sciences An Intenational Jounal @ 0 NSP Natual Sciences Publishing o Point-iseial oelation Analysis of Fuzzy Attibutes Hao-En hueh

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

Annales UMCS Informatica AI 2 (2004) UMCS

Annales UMCS Informatica AI 2 (2004) UMCS Pobane z czasopisma Annales AI- Infomatica http://ai.annales.umcs.pl Annales Infomatica AI 2 (2004) 33-340 Annales Infomatica Lublin-Polonia Sectio AI http://www.annales.umcs.lublin.pl/ Embedding as a

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

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

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

APPLICATION OF STRUCTURED QUEUING NETWORKS IN QOS ESTIMITION OF TELECOMMUNICATION SERVICE

APPLICATION OF STRUCTURED QUEUING NETWORKS IN QOS ESTIMITION OF TELECOMMUNICATION SERVICE APPLICATION OF STRUCTURED QUEUING NETWORKS IN QOS ESTIMITION OF TELECOMMUNICATION SERVICE 1 YAROSLAVTSEV A.F., 2 Al-THUNEIBAT S.A., 3 AL TAWALBEH N.A 1 Depatment of Netwoking, SSUTI, Novosibisk, Russia

More information

Lecture 27: Voronoi Diagrams

Lecture 27: Voronoi Diagrams We say that two points u, v Y ae in the same connected component of Y if thee is a path in R N fom u to v such that all the points along the path ae in the set Y. (Thee ae two connected components in the

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

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

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

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

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

Efficient protection of many-to-one. communications

Efficient protection of many-to-one. communications Efficient potection of many-to-one communications Miklós Molná, Alexande Guitton, Benad Cousin, and Raymond Maie Iisa, Campus de Beaulieu, 35 042 Rennes Cedex, Fance Abstact. The dependability of a netwok

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

Separability and Topology Control of Quasi Unit Disk Graphs

Separability and Topology Control of Quasi Unit Disk Graphs Sepaability and Topology Contol of Quasi Unit Disk Gaphs Jiane Chen, Anxiao(Andew) Jiang, Iyad A. Kanj, Ge Xia, and Fenghui Zhang Dept. of Compute Science, Texas A&M Univ. College Station, TX 7784. {chen,

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

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

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

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

Embeddings into Crossed Cubes

Embeddings into Crossed Cubes Embeddings into Cossed Cubes Emad Abuelub *, Membe, IAENG Abstact- The hypecube paallel achitectue is one of the most popula inteconnection netwoks due to many of its attactive popeties and its suitability

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

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

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

UCLA Papers. Title. Permalink. Authors. Publication Date. Localized Edge Detection in Sensor Fields. https://escholarship.org/uc/item/3fj6g58j

UCLA Papers. Title. Permalink. Authors. Publication Date. Localized Edge Detection in Sensor Fields. https://escholarship.org/uc/item/3fj6g58j UCLA Papes Title Localized Edge Detection in Senso Fields Pemalink https://escholashipog/uc/item/3fj6g58j Authos K Chintalapudi Govindan Publication Date 3-- Pee eviewed escholashipog Poweed by the Califonia

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

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

DISTRIBUTION MIXTURES

DISTRIBUTION MIXTURES Application Example 7 DISTRIBUTION MIXTURES One fequently deals with andom vaiables the distibution of which depends on vaious factos. One example is the distibution of atmospheic paametes such as wind

More information

Root Cause Analysis for Long-Lived TCP Connections

Root Cause Analysis for Long-Lived TCP Connections Root Cause Analysis fo Long-Lived TCP Connections M. Siekkinen, G. Uvoy-Kelle, E. W. Biesack, T. En-Najjay Institut Euécom BP 93, 694 Sophia-Antipolis Cedex, Fance Email: {siekkine,uvoy,ebi,ennajja}@euecom.f

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

(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

Experimental and numerical simulation of the flow over a spillway

Experimental and numerical simulation of the flow over a spillway Euopean Wate 57: 253-260, 2017. 2017 E.W. Publications Expeimental and numeical simulation of the flow ove a spillway A. Seafeim *, L. Avgeis, V. Hissanthou and K. Bellos Depatment of Civil Engineeing,

More information

Input Layer f = 2 f = 0 f = f = 3 1,16 1,1 1,2 1,3 2, ,2 3,3 3,16. f = 1. f = Output Layer

Input Layer f = 2 f = 0 f = f = 3 1,16 1,1 1,2 1,3 2, ,2 3,3 3,16. f = 1. f = Output Layer Using the Gow-And-Pune Netwok to Solve Poblems of Lage Dimensionality B.J. Biedis and T.D. Gedeon School of Compute Science & Engineeing The Univesity of New South Wales Sydney NSW 2052 AUSTRALIA bbiedis@cse.unsw.edu.au

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

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

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

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

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

Hierarchical Region Mean-Based Image Segmentation

Hierarchical Region Mean-Based Image Segmentation Hieachical Region Mean-Based Image Segmentation Slawo Wesolkowski and Paul Fieguth Systems Design Engineeing Univesity of Wateloo Wateloo, Ontaio, Canada, N2L-3G1 s.wesolkowski@ieee.og, pfieguth@uwateloo.ca

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 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

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

Generalized Grey Target Decision Method Based on Decision Makers Indifference Attribute Value Preferences

Generalized Grey Target Decision Method Based on Decision Makers Indifference Attribute Value Preferences Ameican Jounal of ata ining and Knowledge iscovey 27; 2(4): 2-8 http://www.sciencepublishinggoup.com//admkd doi:.648/.admkd.2724.2 Genealized Gey Taget ecision ethod Based on ecision akes Indiffeence Attibute

More information

MapReduce Optimizations and Algorithms 2015 Professor Sasu Tarkoma

MapReduce Optimizations and Algorithms 2015 Professor Sasu Tarkoma apreduce Optimizations and Algoithms 2015 Pofesso Sasu Takoma www.cs.helsinki.fi Optimizations Reduce tasks cannot stat befoe the whole map phase is complete Thus single slow machine can slow down the

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

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

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

Information Retrieval. CS630 Representing and Accessing Digital Information. IR Basics. User Task. Basic IR Processes

Information Retrieval. CS630 Representing and Accessing Digital Information. IR Basics. User Task. Basic IR Processes CS630 Repesenting and Accessing Digital Infomation Infomation Retieval: Basics Thosten Joachims Conell Univesity Infomation Retieval Basics Retieval Models Indexing and Pepocessing Data Stuctues ~ 4 lectues

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

Topic -3 Image Enhancement

Topic -3 Image Enhancement Topic -3 Image Enhancement (Pat 1) DIP: Details Digital Image Pocessing Digital Image Chaacteistics Spatial Spectal Gay-level Histogam DFT DCT Pe-Pocessing Enhancement Restoation Point Pocessing Masking

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

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

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

Also available at ISSN (printed edn.), ISSN (electronic edn.) ARS MATHEMATICA CONTEMPORANEA 3 (2010)

Also available at  ISSN (printed edn.), ISSN (electronic edn.) ARS MATHEMATICA CONTEMPORANEA 3 (2010) Also available at http://amc.imfm.si ISSN 1855-3966 (pinted edn.), ISSN 1855-3974 (electonic edn.) ARS MATHEMATICA CONTEMPORANEA 3 (2010) 109 120 Fulleene patches I Jack E. Gave Syacuse Univesity, Depatment

More information

The Internet Ecosystem and Evolution

The Internet Ecosystem and Evolution The Intenet Ecosystem and Evolution Contents Netwok outing: basics distibuted/centalized, static/dynamic, linkstate/path-vecto inta-domain/inte-domain outing Mapping the sevice model to AS-AS paths valley-fee

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

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

(a, b) x y r. For this problem, is a point in the - coordinate plane and is a positive number.

(a, b) x y r. For this problem, is a point in the - coordinate plane and is a positive number. Illustative G-C Simila cicles Alignments to Content Standads: G-C.A. Task (a, b) x y Fo this poblem, is a point in the - coodinate plane and is a positive numbe. a. Using a tanslation and a dilation, show

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

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

COMPARISON OF CHIRP SCALING AND WAVENUMBER DOMAIN ALGORITHMS FOR AIRBORNE LOW FREQUENCY SAR DATA PROCESSING

COMPARISON OF CHIRP SCALING AND WAVENUMBER DOMAIN ALGORITHMS FOR AIRBORNE LOW FREQUENCY SAR DATA PROCESSING COMPARISON OF CHIRP SCALING AND WAVENUMBER DOMAIN ALGORITHMS FOR AIRBORNE LOW FREQUENCY SAR DATA PROCESSING A. Potsis a, A. Reigbe b, E. Alivisatos a, A. Moeia c,and N. Uzunoglu a a National Technical

More information

Image Enhancement in the Spatial Domain. Spatial Domain

Image Enhancement in the Spatial Domain. Spatial Domain 8-- Spatial Domain Image Enhancement in the Spatial Domain What is spatial domain The space whee all pixels fom an image In spatial domain we can epesent an image by f( whee x and y ae coodinates along

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

A Family of Distributed Deadlock Avoidance Protocols and their Reachable State Spaces

A Family of Distributed Deadlock Avoidance Protocols and their Reachable State Spaces A Family of Distibuted Deadlock Avoidance Potocols and thei Reachable State Spaces Césa Sánchez, Henny B. Sipma, and Zoha Manna Compute Science Depatment Stanfod Univesity, Stanfod, CA 94305-9025 {cesa,sipma,manna}@cs.stanfod.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

Comparisons of Transient Analytical Methods for Determining Hydraulic Conductivity Using Disc Permeameters

Comparisons of Transient Analytical Methods for Determining Hydraulic Conductivity Using Disc Permeameters Compaisons of Tansient Analytical Methods fo Detemining Hydaulic Conductivity Using Disc Pemeametes 1,,3 Cook, F.J. 1 CSRO Land and Wate, ndoooopilly, Queensland The Univesity of Queensland, St Lucia,

More information

Spiral Recognition Methodology and Its Application for Recognition of Chinese Bank Checks

Spiral Recognition Methodology and Its Application for Recognition of Chinese Bank Checks Spial Recognition Methodology and Its Application fo Recognition of Chinese Bank Checks Hanshen Tang 1, Emmanuel Augustin 2, Ching Y. Suen 1, Olivie Baet 2, Mohamed Cheiet 3 1 Cente fo Patten Recognition

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

Parallel processing model for XML parsing

Parallel processing model for XML parsing Recent Reseaches in Communications, Signals and nfomation Technology Paallel pocessing model fo XML pasing ADRANA GEORGEVA Fac. Applied Mathematics and nfomatics Technical Univesity of Sofia, TU-Sofia

More information

Multi-azimuth Prestack Time Migration for General Anisotropic, Weakly Heterogeneous Media - Field Data Examples

Multi-azimuth Prestack Time Migration for General Anisotropic, Weakly Heterogeneous Media - Field Data Examples Multi-azimuth Pestack Time Migation fo Geneal Anisotopic, Weakly Heteogeneous Media - Field Data Examples S. Beaumont* (EOST/PGS) & W. Söllne (PGS) SUMMARY Multi-azimuth data acquisition has shown benefits

More information

UNION FIND. naïve linking link-by-size link-by-rank path compression link-by-rank with path compression context. An Improved Equivalence Algorithm

UNION FIND. naïve linking link-by-size link-by-rank path compression link-by-rank with path compression context. An Improved Equivalence Algorithm Disjoint-sets data type Lectue slides by Kevin Wayne Copyight 5 Peason-Addison Wesley http://www.cs.pinceton.edu/~wayne/kleinbeg-tados UNION FIND naïve linking link-by-size link-by-ank path compession

More information

SYSTEM LEVEL REUSE METRICS FOR OBJECT ORIENTED SOFTWARE : AN ALTERNATIVE APPROACH

SYSTEM LEVEL REUSE METRICS FOR OBJECT ORIENTED SOFTWARE : AN ALTERNATIVE APPROACH I J C A 7(), 202 pp. 49-53 SYSTEM LEVEL REUSE METRICS FOR OBJECT ORIENTED SOFTWARE : AN ALTERNATIVE APPROACH Sushil Goel and 2 Rajesh Vema Associate Pofesso, Depatment of Compute Science, Dyal Singh College,

More information

And Ph.D. Candidate of Computer Science, University of Putra Malaysia 2 Faculty of Computer Science and Information Technology,

And Ph.D. Candidate of Computer Science, University of Putra Malaysia 2 Faculty of Computer Science and Information Technology, (IJCSIS) Intenational Jounal of Compute Science and Infomation Secuity, Efficient Candidacy Reduction Fo Fequent Patten Mining M.H Nadimi-Shahaki 1, Nowati Mustapha 2, Md Nasi B Sulaiman 2, Ali B Mamat

More information

Pipes, connections, channels and multiplexors

Pipes, connections, channels and multiplexors Pipes, connections, channels and multiplexos Fancisco J. Ballesteos ABSTRACT Channels in the style of CSP ae a poeful abstaction. The ae close to pipes and connections used to inteconnect system and netok

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