Power-Aware Mapping for Network-on-Chip Architectures under Bandwidth and Latency Constraints

Size: px
Start display at page:

Download "Power-Aware Mapping for Network-on-Chip Architectures under Bandwidth and Latency Constraints"

Transcription

1 Power-Aware Mappng for Network-on-Chp Archtectures under Bandwdth and Latency Constrants Xaohang Wang 1,2, Me Yang 2, Yngtao Jang 2, and Peng Lu 1 1 Department of Informaton Scence and Electronc Engneerng, Zheang Unversty, Hangzhou, Chna, Department of Electrcal and Computer Engneerng, Unversty of Nevada, Las Vegas, Las Vegas, NV Emals: 1 bakena@yahoo.com.cn, lupeng@see.zu.edu.cn, 2 meyang, yngtao@egr.unlv.edu Abstract Ths paper nvestgates the bandwdth- and latencyconstraned IP mappng problem that maps a gven set of IP cores onto the tles of a mesh-based Network-on-Chp (NoC archtecture to mnmze the power consumpton due to ntercore communcatons. By examnng varous applcatons communcaton characterstcs shown n ther communcaton trace graphs, two dstngushable connectvty templates are realzed: the graphs wth tghtly coupled vertces and those wth dstrbuted vertces. Dfferent mappng heurstcs are developed for these templates: tghtly coupled vertces are mapped onto tles that are close to each other whle the dstrbuted vertces are mapped followng a graph partton scheme. The proposed template-based mappng algorthm acheves on average 15% power savng compared wth MOCA, a fast greedy-based algorthm. Compared wth a branch-and-bound algorthm, the proposed algorthm can generate results of almost the same qualty but requre much less CPU tme. Keywords-Network-on-Chp (NoC; power-aware; IP mappng; Template-based Effcent Mappng (TEM. I. INTRODUCTION Wth the contnuous scalng of CMOS technologes, nterconnects domnate both performance and power dsspaton n future System-on-Chp (SoC desgns. Multprocessor System on Chp (MPSoC [2] desgns have emerged and shown to delver hgh performance yet reasonably low power consumpton. Networks-on-chp (NoC has been consdered as a vable communcaton nfrastructure n MPSoC [9]. Regular mesh remans the domnant NoC archtecture of choce due to ts dstnct features: structured network wrng, modularty, and standard nterfaces [8]. Ths paper focuses on mappng the IP cores onto a regular tle-based NoC archtecture and the mappng s subected to the latency and bandwdth constrants as mposed by many real-tme multmeda applcatons. In the lterature, a number of algorthms have been proposed to solve ths IP mappng problem, and they fall nto four general categores: Branch-and-bound algorthms [8], whch can generate very hgh qualty results. Wth large queue sze, however, ths algorthm demands hgh memory depth and suffers from long CPU tme. Framework-based approaches, such as the ones usng smulated annealng (SA/genetcs algorthm (GA/tabu search (TS [1, 6, 10, 11]. These algorthms typcally requre consderably longer tme than a greedy-based mappng algorthm. Lnear programmng based schemes [13]. Computaton s tme-consumng and there s no guarantee that hghqualty solutons can be always found. Greedy-based heurstcs [5, 12, 15], whch requre sgnfcantly low CPU tme. The followng greedy algorthms are based on dfferent observatons. The algorthm MOCA [15] acheves a sound balance between the run tme and the qualty of solutons. The problem of ths algorthm s that the graph partton algorthm may separate some tghtly coupled regons (.e. the regons made of IP cores wth large degrees and strongly connected nto dfferent sets, makng the vertces n a tghtly coupled regon adversely mapped onto tles physcally far apart. Partcularly, the power performance of MOCA deterorates drastcally when latency constrants are appled. In ths paper, we propose a Template-based Effcent Mappng (TEM algorthm whch generates hgh-qualty mappng results wth low run tme. Ths algorthm s desgned based on two dstngushable connectvty templates extracted from varous applcatons communcaton trace graphs: the graphs wth tghtly coupled vertces and those wth dstrbuted vertces. Correspondngly, dfferent mappng strateges are proposed for these two templates. Smulaton results show that the proposed TEM algorthm acheves better result than MOCA [15]. The rest of the paper s organzed as follows. Secton II formally defnes the mappng problem. Secton III ntroduces the two templates derved from varous applcatons communcaton trace graphs. The mappng algorthm based on the two templates s presented n Secton IV. Secton V presents and dscusses the smulaton results, and fnally, Secton VI concludes the paper. Ths work s n part supported by NSF under grant no. ECCS and NSF of Chna under grant no

2 II. PROBLEM FORMULATION A. Archtecture descrpton and power model Wthout loss of generalty, the NoC system under consderaton s composed of NxN tles nterconnected by a 2- D mesh network. The power model used n [8] s followed n ths study. The average power consumpton for sendng one bt of data from tle t to tle t can be represented as t, t E bt = ηhops ESbt + ( ηhops 1 ELbt (2 where η hops s the number of routers traversed from tle t to tle t, E Sbt s the power consumed by the swtch, and E Lbt s the power consumed on the lnks between tles t and t. B. Problem Descrpton We assume that before IP mappng s performed, a gven applcaton descrbed by a set of concurrent tasks s already bounded and scheduled onto a lst of selected IP cores. As defned below, the communcaton patterns between any par of IP cores of an applcaton are descrbed by ts Communcaton Trace Graph, whereas the NoC archtecture s modeled by ts Archtecture Characterzaton Graph. Defnton 1 A Communcaton Trace Graph (CTG G=(P, E s an undrected graph, where a vertex/node p k P represents an IP core (a processor, an ASIC devce or a memory unt, etc., and an edge e =(p k, p E represents the communcaton trace between vertces p k and p. For each edge e, ω(e defnes the communcaton bandwdth request between vertces p and p gven n bts per second (bps. ω(e sets the mnmum bandwdth that should be allocated by the network n order to meet the performance constrants. σ(e represents the latency constrant, whch s gven n number of hops nstead of an absolute number n cycles [15]. W(e represents the weght of edge e. The weght s defned n the same way as that n [15]. Among all the traces n the graph, let e be the trace wth the hghest bandwdth requrement, and e be the trace wth the tghtest (lowest latency constrant. An nteger K s defned as the mnmum value requred to ensure that ω( e ω( e among all the traces n the graph,. Once K K σ( e σ( e K s determned, each edge s assgned a weght ω( e We= (. σ ( e K Defnton 2 An Archtecture Characterzaton Graph (ACG Ğ=(T, L s an undrected graph, where each vertex t T represents a tle and each edge l L=(t k, t represents the lnk between tles t k and t. For each lnk l, bw(l defnes the bandwdth provded between tles t k and t. c(l defnes the lnk cost of l,.e., power consumpton for transmttng one bt data from t k to t. In ths paper, we focus on regular NoC archtectures whch have bw(l =B, c(l = C for each l L, where B and C are constants. h k, s the set of lnks formng one of the shortest paths from tle t k to tle t (h k, L. dst(h k, determnes the number of elements n h k, (.e. t s the hop count of the shortest path between tle t k and tle t. Defnton 3 A mappng algorthm M: P T maps each vertex n CTG onto an avalable tle n ACG. M(p represents the mapped tle n ACG, where p P and M(p T. Defnton 4 A routng algorthm R: E H, fnds one of the shortest routng path between M(p k and M(p for each edge e =(p k, p E. The lnks of formng ths path belongs to set h M ( pk, M( p. The IP mappng problem s formulated as follows. Gven a CTG(P, E representng the communcaton pattern of an applcaton and an ACG(T, L representng the target NoC archtecture, where P T, fnd a mappng M:P T whch maps all the vertces n CTG onto avalable tles n ACG and generates a deadlock-free and mnmal routng paths for all edges n CTG, such that the total power consumpton s mnmzed,.e., satsfyng E hm ( p, M ( p k (3 Mn ω( e C = 0 m= 0 e= ( pk, p E lm hm ( pk, M ( p p P, M( p, p P and p T, (4 p p, M( p M( p, (5 ω( e f( l, hm p M p (6 l, B (, ( e= ( pk, p e =(p k, p, σ ( e dst ( h M( p, ( k M p (7 1 f l h M( pk, M( p where f( l, h M( p, ( k M p = 0 f l h M ( pk, M ( p Smlar to the defnton adopted n [8], condtons gven by (4 and (5 ensure that each IP should be mapped exactly to one tle and no tle can host more than one IP. Eqn. (6 specfes the bandwdth constrant for every lnk set by the communcaton bandwdth requrement, and Eqn. (7 ensures that the latency constrant (n terms of the number of hops between two communcatng IPs s satsfed on the mapped tles. III. DERIVATION OF MAPPING TEMPLATES As alluded n Secton 1, MOCA [15] does not perform well when the latency constrants are consdered. Ths problem s llustrated n the mappng of the MPEG4 decoder, wth ts CTG shown n Fgure 1. k

3 Fgure 1 The CTG of MPEG4 decoder [15]. The BW request (Mbps and latency constrant (number of hops of each edge s labeled on the edge. A communcaton path wth hgher bandwdth request should be mapped to lnks wth fewer hop counts to reduce power consumpton as Eqn. (3 ndcates. When the latency constrant s appled, the result of MOCA deterorates. For example, n Fgure 2(a, the hop counts of edges (1,7, (8,9, (8,10 are greater than 1, whch results n hgher power consumpton. As a matter of fact, as shown n Fgure 2(b, there s a better mappng by reducng the number of hops for these edges. In Fgure 1, vertces 1 and 8 have many edges requestng hgh bandwdths or tght latency. In Fgure 2(b, these two vertces are mapped onto the two tles wth the largest degree or maxmal number of neghbor tles (.e., 4. In ths soluton, edges wth hgher weghts are mapped wth fewer hop counts than those edges wth smaller weghts. (a (b Fgure 2 (a The mappng result of MOCA on the CTG of MPEG4dec w/ latency constrants (b A better mappng soluton on the same CTG w/ latency constrant. The square boxes are tles and the crcles are IP cores. From the above example, one can see that dfferent mappng strateges shall be adopted for CTGs wth dfferent features. Here, we present two dstnct templates derved from varous CTGs. Gven a CTG(P, E, and a sorted lst of edges n decreasng order of edge weght denoted as Ê: Defnton 5 A vertex p P s a hot node f a p has a degree greater than or equal to 4, and b of the frst 50% edges n the edge lst Ê, there are at least one edges that are connected to p. Template 1: An applcaton s CTG falls nto Template 1 (tghtly coupled f there s at least one hot node n the CTG. Template 2: An applcaton falls nto Template 2 (dstrbuted f there s no hot node n the applcaton s CTG. IV. ALGORITHM DESCRIPTION The overall structure of the TEM algorthm s shown below. Before the template-based mappng takes place, the edges have to be sorted n non-ncreasng order n terms of the edge weght. After all the vertces are mapped, a routng allocaton routne s called to fnd routng paths for all pars of communcatng vertces. TEM (G(P,E, Ğ(T,L Input: (1 G: CTG of an applcaton. (2 Ğ: ACG of an NoC archtecture Functon: Map the applcaton on the NoC archtecture and allocate routng paths Sort_Edge(E; // E G. Let the sorted lst be Ê =e π(1, e π(2,,e π( E,.e., // W(e π(1 W(e π(2 W(e π(( E. // Fnd the hot nodes f any. Check whch template the applcaton belongs to for each vertex p // Defnton 5 f (vertex p has degree α and p has γ edges n the frst β edges n E mark p as a hot node an place nto set Ĥ // If G has hot nodes, t belongs to Template 1, otherwse Template 2 // Template specfc algorthm case 1: Template 1 // Tghtly coupled TEM_Template1(G, Ğ, Ĥ; break; case 2: Template 2 // Dstrbuted TEM_Template2(G, Ğ; break; Route_Alloc(G, Ğ; // Fnd routng paths A. Mappng Algorthm for Template 1 An applcaton of template 1 (CTG has at least one hot node s mapped based on the followng observatons. a Hot nodes should be gven a hgher mappng prorty; that s, they shall be mapped before any other nodes are mapped. All the hot nodes n a CTG wll be frst mapped along wth ther α most sgnfcant neghbor vertces. A hot node s better mapped onto a tle n an NoC that has the maxmum number of neghbor tles. b Once all the hot nodes are mapped, the mappng sequence of remanng unmapped non-hot nodes wll be performed based on the decreasng order of weght of edges connectng them. As such, TEM_Template1 procedure conssts of two maor steps: (1 map hot nodes n CTG, (2 map other vertces. TEM_Template1 (G(P,E, Ğ(T,L, Ĥ Input: (1 G: CTG of an applcaton (2 Ğ: ACG of an NoC archtecture (3 Ĥ: The set of hot nodes Functon: Map the Template 1 applcaton onto the NoC archtecture // Step 1: fnd and map the hot nodes wth ther 4 most sgnfcant // neghbors for each edge e π(=(p k, p ncrease the counter of p k,, p f they are hot nodes (belong to Ĥ f one of the vertex s counter equals α // suppose p k s counter s α Map_Hot_Node(p k, Ê, T; // Step 2: map the remanng unmapped vertces update the hop counts of edges whose termnal vertces are already mapped; // re-scan the edge lst for each edge e π(=(p k, p Map_Edge(e π(, Ê, T; // T s the avalable tle set

4 Before we dscuss each step n detal, the followng defntons are ntroduced. Defnton 6 Vertex p n s a sgnfcant neghbor of vertex p f there exsts an edge (p, p n E and edge (p, p n s wthn the frst 50% edges n the edge lst Ê. Defnton 7 Snce n a mesh structure, the maxmal degree of a tle s four, here the four most sgnfcant neghbors p n1,, p n4 of a hot node p are the four neghbors of p that have the hghest bandwdth/tghtest latency requrements among all the neghbors of p. Defnton 8 A center tle n a 2-D topology lke mesh s a tle that has maxmal number (.e. 4 of unmapped neghbor tles. Defnton 9 Two vertces p k, p are close f (1 edge (p k, p s among the frst 50% edges n the edge lst or (2 a neghbor vertex of p k (or p s connected wth p (or p k, and ths edge s among the frst 50% edges n the edge lst. 1 Map hot nodes. Procedure Map_Hot_Node maps each hot node and ts four most sgnfcant neghbor vertces. Each hot node p s assocated wth a counter ctr that counts the number of tmes that vertex p appears as a termnal vertex of the edges n the sorted edge lst. Case 1: the current hot node p s not mapped yet. Check whether ths hot node s close (Defnton 9 to any of those mapped hot nodes. If yes, ths hot node s mapped to a tle wth mnmum hops to an already mapped hot node. Otherwse, a center tle s selected and t s allocated to ths hot node. Next ts four most sgnfcant neghbors are mapped to the neghborng tles f they have not been mapped. Case 2: the current hot node has already been mapped because t belongs to the frst α neghbors of a prevously mapped hot node. In ths case, only the four most sgnfcant neghbor vertces need to be mapped. The procedure Improve_Edge can be called for optmzaton. 2 Map other nodes. In ths step, all the remanng unmapped vertces are mapped. Procedure Map_Edge maps the remanng unmapped vertex/vertces. There are three cases to consder. Case 1: Nether of the two termnal vertces p k, p of edge e π( = (p k, p s mapped. Search the edges from e π(+1 to e π( E. If one of the neghbors of the two termnal vertces s found and s mapped, then the vertex wth a mapped neghbor s mapped onto a tle that has the mnmum hop count to ts mapped neghbor. On the other hand, f none of the two termnal vertces neghbors s mapped, an avalable tle s selected and mmedately allocated to one of the vertces, after whch the other vertex s mapped onto a tle wth a mnmum hop count to the ust mapped tle. Case 2: One of the two termnal vertces s mapped, but the other one s not. In ths case, only the unmapped vertex needs to be mapped onto a tle wth mnmum hop count to the mapped tle. The Map_Hot_Node procedure s lsted below Map_Hot_Node(p h, Ê, T Input: (1 p h: A hot node (2 Ê: The sorted edge lst (3 T: The avalable tle set n NoC archtecture Functon: Map hot node p h and ts neghbors // Case 1 f (p h s unmapped // 1 Map the hot node p h check f other mapped hot nodes are close to p h ; // Defnton 9 f (there s a mapped hot node p hk close to hot node p h map p h to a tle wth mnmum hop count to p h s tle; else map p h to t l = Fnd_Center_Tle(; // 2 Map the four most sgnfcant neghbors for each of p h s four most sgnfcant neghbors // suppose p n s one of such neghbors f (p n s unmapped // Two crtera should be observed n mappng neghbors as // descrbed n text map p n to a tle wth mnmum hops to p h s tle; // Case 2 else f (p h s mapped for each of p h s four most sgnfcant neghbors // suppose p n s one of such neghbors f (p n s not mapped map p n to a tle wth mnmum hop count to p h s tle ; Fgure 2(b shows the mappng result of the CTG of MPEG4 decoder. The complexty of TEM_Template1 s O( Ĥ ( E + T + E ( E + T, whch can be further smplfed as O(( E + T 2 snce Ĥ P T. B. TEM Algorthm for Template 2 For a Template 2 CTG, the graph partton algorthm proposed n [4] s followed to partton both CTG and ACG nto four smaller regons so that each regon n CTG can be mapped onto a regon n ACG n a dvde-and-conquer manner. Defnton 10 A Block Trace Graph (BTG s denoted as G'=(B, BE, where each vertex b B s a parttoned block and b P (P s the vertex set of CTG(P, E, and an edge be =(b k, b BE exsts f p kl b k and p m b, (p kl, p m E. For an edge be =(b k, b BE, W(be represents the weght of be. The calculaton s the same as the weght of edges n CTG. An CTG edge (p, p E belongs to a block b k f p b k or p b k. E represents the set of edges belongng to b. The vertces n each block are classfed nto two types and they are mapped dfferently: 1 Internal vertces. An nternal vertex p ntern has all ts neghbors n the same block of p ntern. 2 External vertces. An external vertex p extern has at least one of ts neghbors not n the same block of p extern. Defnton 11 A Vrtual ACG (VACG s denoted as Ğ'=(R, CH, where a vertex r R represents a parttoned regon and r T (T s the tle set of ACG(T, L. Each edge ch =(r k,

5 r CH represents that there exsts drect lnks between tles n r k and r. For each edge ch CH, The TEM_Template 2 algorthm works as follows. TEM_Template2(G(P,E, Ğ(T,L Input: (1 G: The CTG of an applcaton. (2 Ğ: The ACG of an NoC archtecture Functon: Map the Template 2 applcaton onto the NoC archtecture // Step 1. Obtan the BTG and VACG by parttonng the CTG nto blocks // and ACG nto default regons, respectvely. Map the blocks n BTG to // the regons n VACG. obtan BTG (G' usng the graph partton algorthm to partton CTG nto 4 blocks; obtan VACG (Ğ' wth default four square regons; map the four blocks n BTG to four default regons of VACG usng Map_Edge; Partton_VACG(G', Ğ'; // Step 2. Insde each block, map the vertces to the tles of the // correspondng regon. for each block b n BTG mapped to regon r x n VACG // 1 Map the external vertces. // Let PE be set of external vertces n b ; sort external vertces n PE n the non-ncreasng order of the total weght of the external vertces. Let p Φ(1, p Φ (2,, p Φ ( PE be the sorted lst; for each external vertex p Φ( Map_External_Vertex(p Φ(, G', Ğ'; // 2 Map the nternal vertces sort the edges n b nto Ê usng Sort_Edge; for each edge e ψ ( = (p k, p Ê Map_Edge (e ψ(, b, r x; // descrbed n Secton Partton. In ths step, the BTG s formed usng a graph partton algorthm [4] to partton the CTG nto four blocks. The VACG s formed by parttonng the ACG nto four default regons (e.g., square regons n Fgure 4. The blocks n the BTG are mapped to the default regons n the VACG. Based on the mappng result of BTG onto VACG, the actual regons n the VACG are parttoned n a top-down manner. The sze of each parttoned regon needs to be set equal to the sze of the correspondng block n the BTG. 2 Map nsde blocks. In ths step, the vertces n each block are mapped to the tles wthn ts regon. The external vertces are frst mapped to the border of each NoC regon, after whch the nternal vertces are mapped wth larger weght edge frst crteron. The overall complexty of TEM_Template2 can be approxmated by O(( E + T 2 as blk < P T, E < E, T < T. V. PERFORMANCE EVALUATION To evaluate the performance of the TEM algorthm, both TEM_Template 1 and TEM_Template2 are mplemented and smulated. The brand-and-bound (BNB [8] and MOCA [15] algorthms are also mplemented and the mappng results of all three algorthms are compared. To solate the effects of dfferent routng algorthms, the XY routng and odd-even routng algorthms are selected. Multmeda benchmarks from [7] and [15]. The Noxm smulator [14] s modfed and used n our smulatons to obtan the total communcaton power after the mappng and the routng path allocaton steps. The smulatons are performed on a PC wth one Intel Core2 P GHz processor and 2GB RAM. The run tmes of both TEM and MOCA range from 0.01sec~1sec for dfferent network szes. Multmeda benchmarks of both Template 1 and Template 2 are tested on a mesh-based NoC. Table I lsts these benchmarks. For benchmarks of both templates, we compare the degradaton of TEM compared to BNB (.e. the ncrease n power consumpton of mappng results from the TEM compared to that from BNB and the degradaton of MOCA compared to BNB. The power consumpton s normalzed. TABLE I. MULTIMEDIA BENCHMARKS AND THEIR TEMPLATES. Benchmark Template MPEG enc 1 263enc+MP3enc 1 263enc+MP3dec 1 263enc+263dec 1 Multmeda Systems (MMS 1 VOPD 2 MP3enc 2 MP3enc+MP3dec 2 Table II shows the result of benchmarks of Template 1 wth and wthout latency constrants. When the latency constrants are not consdered, the result of TEM s comparable wth that of MOCA. When the latency constrants are consdered, the TEM algorthm outperforms MOCA for most meda programs. The degradaton of TEM vs. BNB s wthn 10% wth odd-even routng. The reducton n power consumpton of TEM compared to MOCA s over 15%. For MPEG4dec, the degradaton of TEM s only 6% compared to that of MOCA 45%. The Mult-Meda System (MMS benchmark [8] s also smulated on a 5x5 mesh-based NoC. Table III shows the reducton n power consumpton of TEM over MOCA (normalzed. Table IV shows the result of benchmarks of Template 2 wth and wthout latency constrants. Wthout latency constrants, the result of TEM s comparable wth that of MOCA. On average, the degradaton of TEM over BNB s wthn 10% for both XY routng and odd-even routng. Wth latency constrants, the power consumpton of mappng result from TEM s slghtly lower than that from MOCA for VOPD and MP3enc. For MP3enc+MP3dec, TEM acheves sgnfcant reducton n power consumpton compared to MOCA. Ths s due to the fact that the TEM algorthm for applcatons of Template 2 uses a dvde-and-conquer approach. Insde each parttoned block, TEM frst maps external vertces to the border of the block and the remanng edges wth larger weght frst crteron. Thus larger weght edges are mapped frst nsde each block. Evaluaton of TEM on random applcatons s also conducted. Detaled results are presented n [16].

6 TABLE II. COMPARISON OF POWER CONSUMPTION (NORMALIZED OF TEM AND MOCA OVER BNB ON BENCHMARKS OF TEMPLATE 1. Benchmarks Wth latency constrant Wthout latency constrant TEM MOCA BNB TEM MOCA BNB XY Odd-even XY Odd-even MPEG enc enc+MP3enc enc+MP3dec enc+263dec TABLE III. COMPARISON OF POWER CONSUMPTION (NORMALIZED OF TEM OVER MOCA ON MMS (TEMPLATE 1. Benchmarks Wth latency constrant Wthout latency constrant TEM MOCA TEM MOCA XY Odd-even XY Odd-even MMS TABLE IV. COMPARISON OF POWER CONSUMPTION (NORMALIZED TEM AND MOCA OVER BNB ON BENCHMARKS OF TEMPLATE 2. Benchmarks Wth latency constrant Wthout latency constrant TEM MOCA BNB TEM MOCA BNB XY Odd-even XY Odd-even VOPD MP3enc MP3enc+MP3dec VI. CONCLUSION Ths paper presented a template-based greedy algorthm to address the IP mappng problem under the bandwdth and latency constrants. An applcaton falls nto Template 1, f there are one or more hot nodes whch demand ether hgher communcatons bandwdths or tghter latency. In ths case, TEM maps the hot nodes frst along wth ther four most sgnfcant neghbors, after whch the remanng IP cores are mapped n descendng order based on the edge weghts. On the other hand, an applcaton s categorzed as Template 2, f the communcatons are nearly evenly dstrbuted among the vertces. In ths case, TEM dvdes the NoC nto regons and CTG nto blocks; then t maps the IP cores nsde each block n a dvde-and-conquer manner. The experments on multmeda and random benchmarks show that TEM generates hgh qualty mappng results wth low run tmes. VII. REFERENCES [1] G. Asca, V. Catana, and M. Pales, "Mult-obectve mappng for meshbased NoC archtectures," n Proc. 2nd IEEE/ACM/IFIP Int'l Conf Hardware/Software Codesgn and System Synthess, 2004, pp [2] D. Bertozz, A. Jalabert, S. Mural, R. Tamhankar, S. Stergou, L. Benn, and G. De Mchel, "NoC synthess flow for customzed doman specfc multprocessor Systems-on-Chp," IEEE Trans. Parallel And Dstrbuted Systems, pp , [3] R. P. Dck, D. L. Rhodes, and W. Wolf, "TGFF: task graphs for free," n Proc. 6th Int'l workshop on Hardware/software Codesgn, 1998, pp [4] C. M. Fducca and R. M. Mattheyses, "A Lnear-tme heurstc for mprovng network parttons," n Proc. 19th IEEE Conf. Desgn Automaton, 1982, pp [5] A. Hansson, K. Goossens, and A. Rǎ dulescu, "A unfed approach to constraned mappng and routng on network-on-chp archtectures," n Proc. 3rd IEEE/ACM/IFIP Int'l Conf Hardware/Software Codesgn and System Synthess, 2005, pp [6] H. M. Harmanan and R. Farah, "A method for effcent mappng and relable routng for NoC archtectures wth mnmum bandwdth and area," n Proc. Conf. Crcuts and Systems and TAISA, 2008, pp [7] J. Hu and R. Marculescu, "Energy-aware mappng for tle-based NoC archtectures under performance constrants," n Proc Conf. Asa South Pacfc Desgn Automaton, 2003, pp [8] J. Hu and R. Marculescu, "Energy-and performance-aware mappng for regular NoC archtectures," IEEE Trans. Computer-Aded Desgn of Integrated Crcuts and Systems, vol. 24, pp , [9] A. Jantsch and H. Tenhunen, Networks on Chp. New York, NY: Kluwer Academc Publshers, [10] Z. Lu, L. Xa, and A. Jantsch, "Cluster-based smulated annealng for mappng cores onto 2D mesh networks on chp," n Proc. Desgn and Dagnostcs of Electronc Crcuts and Systems (DDECS, 2008, pp [11] C. A. M. Marcon, E. I. Moreno, N. L. V. Calazans, and F. G. Moraes, "Evaluaton of algorthms for low energy mappng onto NoCs," n Proc. IEEE Int'l Symp. Crcuts and Systems(ISCAS, 2007, pp [12] A. Mehran, S. Saed, A. Khademzadeh, and A. Afzal-Kusha, "Spral: a heurstc mappng algorthm for network on chp," J. IEICE Electroncs Express, vol. 4, pp , [13] S. Mural and G. De Mchel, "Bandwdth-constraned mappng of cores onto NoC archtectures," n Proc. Desgn, Automaton and Test n Europe Conference and Exhbton (DATE, 2004, pp [14] Noxm, Avalable at: [15] K. Srnvasan and K. S. Chatha, "A technque for low energy mappng and routng n network-on-chp archtectures," n Proc Int'l Symp. Low Power Electroncs and Desgn, 2005, pp [16] X. Wang, M. Yang, Y. Jang and P. Lu, "A power-aware mappng approach to map IP cores onto NoCs under bandwdth and latency constrants," to appear n ACM Trans. Archtecture and Code Optmzaton.

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

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

More information

An Optimal Algorithm for Prufer Codes *

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

More information

A Fast Content-Based Multimedia Retrieval Technique Using Compressed Data

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

More information

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

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

More information

Routing in Degree-constrained FSO Mesh Networks

Routing in Degree-constrained FSO Mesh Networks Internatonal Journal of Hybrd Informaton Technology Vol., No., Aprl, 009 Routng n Degree-constraned FSO Mesh Networks Zpng Hu, Pramode Verma, and James Sluss Jr. School of Electrcal & Computer Engneerng

More information

Learning the Kernel Parameters in Kernel Minimum Distance Classifier

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

More information

Parallelism for Nested Loops with Non-uniform and Flow Dependences

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

More information

CHAPTER 2 DECOMPOSITION OF GRAPHS

CHAPTER 2 DECOMPOSITION OF GRAPHS CHAPTER DECOMPOSITION OF GRAPHS. INTRODUCTION A graph H s called a Supersubdvson of a graph G f H s obtaned from G by replacng every edge uv of G by a bpartte graph,m (m may vary for each edge by dentfyng

More information

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

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

More information

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

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

More information

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

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

More information

Problem Definitions and Evaluation Criteria for Computational Expensive Optimization

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

More information

6.854 Advanced Algorithms Petar Maymounkov Problem Set 11 (November 23, 2005) With: Benjamin Rossman, Oren Weimann, and Pouya Kheradpour

6.854 Advanced Algorithms Petar Maymounkov Problem Set 11 (November 23, 2005) With: Benjamin Rossman, Oren Weimann, and Pouya Kheradpour 6.854 Advanced Algorthms Petar Maymounkov Problem Set 11 (November 23, 2005) Wth: Benjamn Rossman, Oren Wemann, and Pouya Kheradpour Problem 1. We reduce vertex cover to MAX-SAT wth weghts, such that the

More information

Simulation Based Analysis of FAST TCP using OMNET++

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

More information

Chapter 1. Introduction

Chapter 1. Introduction Chapter 1 Introducton 1.1 Parallel Processng There s a contnual demand for greater computatonal speed from a computer system than s currently possble (.e. sequental systems). Areas need great computatonal

More information

Module Management Tool in Software Development Organizations

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

More information

Constructing Minimum Connected Dominating Set: Algorithmic approach

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

More information

Course Introduction. Algorithm 8/31/2017. COSC 320 Advanced Data Structures and Algorithms. COSC 320 Advanced Data Structures and Algorithms

Course Introduction. Algorithm 8/31/2017. COSC 320 Advanced Data Structures and Algorithms. COSC 320 Advanced Data Structures and Algorithms Course Introducton Course Topcs Exams, abs, Proects A quc loo at a few algorthms 1 Advanced Data Structures and Algorthms Descrpton: We are gong to dscuss algorthm complexty analyss, algorthm desgn technques

More information

Subspace clustering. Clustering. Fundamental to all clustering techniques is the choice of distance measure between data points;

Subspace clustering. Clustering. Fundamental to all clustering techniques is the choice of distance measure between data points; Subspace clusterng Clusterng Fundamental to all clusterng technques s the choce of dstance measure between data ponts; D q ( ) ( ) 2 x x = x x, j k = 1 k jk Squared Eucldean dstance Assumpton: All features

More information

A Load-balancing and Energy-aware Clustering Algorithm in Wireless Ad-hoc Networks

A Load-balancing and Energy-aware Clustering Algorithm in Wireless Ad-hoc Networks A Load-balancng and Energy-aware Clusterng Algorthm n Wreless Ad-hoc Networks Wang Jn, Shu Le, Jnsung Cho, Young-Koo Lee, Sungyoung Lee, Yonl Zhong Department of Computer Engneerng Kyung Hee Unversty,

More information

NUMERICAL SOLVING OPTIMAL CONTROL PROBLEMS BY THE METHOD OF VARIATIONS

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

More information

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

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

More information

Repeater Insertion for Two-Terminal Nets in Three-Dimensional Integrated Circuits

Repeater Insertion for Two-Terminal Nets in Three-Dimensional Integrated Circuits Repeater Inserton for Two-Termnal Nets n Three-Dmensonal Integrated Crcuts Hu Xu, Vasls F. Pavlds, and Govann De Mchel LSI - EPFL, CH-5, Swtzerland, {hu.xu,vasleos.pavlds,govann.demchel}@epfl.ch Abstract.

More information

S1 Note. Basis functions.

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

More information

Routing on Switch Matrix Multi-FPGA Systems

Routing on Switch Matrix Multi-FPGA Systems Routng on Swtch Matrx Mult-FPGA Systems Abdel Enou and N. Ranganathan Center for Mcroelectroncs Research Department of Computer Scence and Engneerng Unversty of South Florda Tampa, FL 33620 Abstract In

More information

An Application of the Dulmage-Mendelsohn Decomposition to Sparse Null Space Bases of Full Row Rank Matrices

An Application of the Dulmage-Mendelsohn Decomposition to Sparse Null Space Bases of Full Row Rank Matrices Internatonal Mathematcal Forum, Vol 7, 2012, no 52, 2549-2554 An Applcaton of the Dulmage-Mendelsohn Decomposton to Sparse Null Space Bases of Full Row Rank Matrces Mostafa Khorramzadeh Department of Mathematcal

More information

Load Balancing for Hex-Cell Interconnection Network

Load Balancing for Hex-Cell Interconnection Network Int. J. Communcatons, Network and System Scences,,, - Publshed Onlne Aprl n ScRes. http://www.scrp.org/journal/jcns http://dx.do.org/./jcns.. Load Balancng for Hex-Cell Interconnecton Network Saher Manaseer,

More information

TPL-Aware Displacement-driven Detailed Placement Refinement with Coloring Constraints

TPL-Aware Displacement-driven Detailed Placement Refinement with Coloring Constraints TPL-ware Dsplacement-drven Detaled Placement Refnement wth Colorng Constrants Tao Ln Iowa State Unversty tln@astate.edu Chrs Chu Iowa State Unversty cnchu@astate.edu BSTRCT To mnmze the effect of process

More information

F Geometric Mean Graphs

F Geometric Mean Graphs Avalable at http://pvamu.edu/aam Appl. Appl. Math. ISSN: 1932-9466 Vol. 10, Issue 2 (December 2015), pp. 937-952 Applcatons and Appled Mathematcs: An Internatonal Journal (AAM) F Geometrc Mean Graphs A.

More information

Non-Split Restrained Dominating Set of an Interval Graph Using an Algorithm

Non-Split Restrained Dominating Set of an Interval Graph Using an Algorithm Internatonal Journal of Advancements n Research & Technology, Volume, Issue, July- ISS - on-splt Restraned Domnatng Set of an Interval Graph Usng an Algorthm ABSTRACT Dr.A.Sudhakaraah *, E. Gnana Deepka,

More information

Sum of Linear and Fractional Multiobjective Programming Problem under Fuzzy Rules Constraints

Sum of Linear and Fractional Multiobjective Programming Problem under Fuzzy Rules Constraints Australan Journal of Basc and Appled Scences, 2(4): 1204-1208, 2008 ISSN 1991-8178 Sum of Lnear and Fractonal Multobjectve Programmng Problem under Fuzzy Rules Constrants 1 2 Sanjay Jan and Kalash Lachhwan

More information

Quality Improvement Algorithm for Tetrahedral Mesh Based on Optimal Delaunay Triangulation

Quality Improvement Algorithm for Tetrahedral Mesh Based on Optimal Delaunay Triangulation Intellgent Informaton Management, 013, 5, 191-195 Publshed Onlne November 013 (http://www.scrp.org/journal/m) http://dx.do.org/10.36/m.013.5601 Qualty Improvement Algorthm for Tetrahedral Mesh Based on

More information

Routability Driven Modification Method of Monotonic Via Assignment for 2-layer Ball Grid Array Packages

Routability Driven Modification Method of Monotonic Via Assignment for 2-layer Ball Grid Array Packages Routablty Drven Modfcaton Method of Monotonc Va Assgnment for 2-layer Ball Grd Array Pacages Yoch Tomoa Atsush Taahash Department of Communcatons and Integrated Systems, Toyo Insttute of Technology 2 12

More information

Efficient Distributed File System (EDFS)

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

More information

Explicit Formulas and Efficient Algorithm for Moment Computation of Coupled RC Trees with Lumped and Distributed Elements

Explicit Formulas and Efficient Algorithm for Moment Computation of Coupled RC Trees with Lumped and Distributed Elements Explct Formulas and Effcent Algorthm for Moment Computaton of Coupled RC Trees wth Lumped and Dstrbuted Elements Qngan Yu and Ernest S.Kuh Electroncs Research Lab. Unv. of Calforna at Berkeley Berkeley

More information

Distributed Middlebox Placement Based on Potential Game

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

More information

Parallel matrix-vector multiplication

Parallel matrix-vector multiplication Appendx A Parallel matrx-vector multplcaton The reduced transton matrx of the three-dmensonal cage model for gel electrophoress, descrbed n secton 3.2, becomes excessvely large for polymer lengths more

More information

Hermite Splines in Lie Groups as Products of Geodesics

Hermite Splines in Lie Groups as Products of Geodesics Hermte Splnes n Le Groups as Products of Geodescs Ethan Eade Updated May 28, 2017 1 Introducton 1.1 Goal Ths document defnes a curve n the Le group G parametrzed by tme and by structural parameters n the

More information

The Codesign Challenge

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

More information

Obstacle-Aware Routing Problem in. a Rectangular Mesh Network

Obstacle-Aware Routing Problem in. a Rectangular Mesh Network Appled Mathematcal Scences, Vol. 9, 015, no. 14, 653-663 HIKARI Ltd, www.m-hkar.com http://dx.do.org/10.1988/ams.015.411911 Obstacle-Aware Routng Problem n a Rectangular Mesh Network Norazah Adzhar Department

More information

A Binarization Algorithm specialized on Document Images and Photos

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

More information

Clustering on antimatroids and convex geometries

Clustering on antimatroids and convex geometries Clusterng on antmatrods and convex geometres YULIA KEMPNER 1, ILYA MUCNIK 2 1 Department of Computer cence olon Academc Insttute of Technology 52 Golomb tr., P.O. Box 305, olon 58102 IRAEL 2 Department

More information

An efficient iterative source routing algorithm

An efficient iterative source routing algorithm An effcent teratve source routng algorthm Gang Cheng Ye Tan Nrwan Ansar Advanced Networng Lab Department of Electrcal Computer Engneerng New Jersey Insttute of Technology Newar NJ 7 {gc yt Ansar}@ntedu

More information

3. CR parameters and Multi-Objective Fitness Function

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

More information

Programming in Fortran 90 : 2017/2018

Programming in Fortran 90 : 2017/2018 Programmng n Fortran 90 : 2017/2018 Programmng n Fortran 90 : 2017/2018 Exercse 1 : Evaluaton of functon dependng on nput Wrte a program who evaluate the functon f (x,y) for any two user specfed values

More information

2x x l. Module 3: Element Properties Lecture 4: Lagrange and Serendipity Elements

2x x l. Module 3: Element Properties Lecture 4: Lagrange and Serendipity Elements Module 3: Element Propertes Lecture : Lagrange and Serendpty Elements 5 In last lecture note, the nterpolaton functons are derved on the bass of assumed polynomal from Pascal s trangle for the fled varable.

More information

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

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

More information

CSE 326: Data Structures Quicksort Comparison Sorting Bound

CSE 326: Data Structures Quicksort Comparison Sorting Bound CSE 326: Data Structures Qucksort Comparson Sortng Bound Steve Setz Wnter 2009 Qucksort Qucksort uses a dvde and conquer strategy, but does not requre the O(N) extra space that MergeSort does. Here s the

More information

Support Vector Machines

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

More information

Storage Binding in RTL synthesis

Storage Binding in RTL synthesis Storage Bndng n RTL synthess Pe Zhang Danel D. Gajsk Techncal Report ICS-0-37 August 0th, 200 Center for Embedded Computer Systems Department of Informaton and Computer Scence Unersty of Calforna, Irne

More information

GSLM Operations Research II Fall 13/14

GSLM Operations Research II Fall 13/14 GSLM 58 Operatons Research II Fall /4 6. Separable Programmng Consder a general NLP mn f(x) s.t. g j (x) b j j =. m. Defnton 6.. The NLP s a separable program f ts objectve functon and all constrants are

More information

Needed Information to do Allocation

Needed Information to do Allocation Complexty n the Database Allocaton Desgn Must tae relatonshp between fragments nto account Cost of ntegrty enforcements Constrants on response-tme, storage, and processng capablty Needed Informaton to

More information

Comparison of Heuristics for Scheduling Independent Tasks on Heterogeneous Distributed Environments

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

More information

Lobachevsky State University of Nizhni Novgorod. Polyhedron. Quick Start Guide

Lobachevsky State University of Nizhni Novgorod. Polyhedron. Quick Start Guide Lobachevsky State Unversty of Nzhn Novgorod Polyhedron Quck Start Gude Nzhn Novgorod 2016 Contents Specfcaton of Polyhedron software... 3 Theoretcal background... 4 1. Interface of Polyhedron... 6 1.1.

More information

AADL : about scheduling analysis

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

More information

A HEURISTIC METHOD FOR RELIABILITY REDUNDANCY OPTIMIZATION OF FLOW NETWORKS

A HEURISTIC METHOD FOR RELIABILITY REDUNDANCY OPTIMIZATION OF FLOW NETWORKS Kumar Pardeep and Chaturved D.K., Pahua G.L. - A HEURISTIC METHOD FOR RELIABILITY REDUNDANCY OPTIMIZATION OF FLOW NETWORKS A HEURISTIC METHOD FOR RELIABILITY REDUNDANCY OPTIMIZATION OF FLOW NETWORKS Kumar

More information

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

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

More information

Goals and Approach Type of Resources Allocation Models Shared Non-shared Not in this Lecture In this Lecture

Goals and Approach Type of Resources Allocation Models Shared Non-shared Not in this Lecture In this Lecture Goals and Approach CS 194: Dstrbuted Systems Resource Allocaton Goal: acheve predcable performances Three steps: 1) Estmate applcaton s resource needs (not n ths lecture) 2) Admsson control 3) Resource

More information

Problem Set 3 Solutions

Problem Set 3 Solutions Introducton to Algorthms October 4, 2002 Massachusetts Insttute of Technology 6046J/18410J Professors Erk Demane and Shaf Goldwasser Handout 14 Problem Set 3 Solutons (Exercses were not to be turned n,

More information

A New Token Allocation Algorithm for TCP Traffic in Diffserv Network

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

More information

Chapter 6 Programmng the fnte element method Inow turn to the man subject of ths book: The mplementaton of the fnte element algorthm n computer programs. In order to make my dscusson as straghtforward

More information

Sequential search. Building Java Programs Chapter 13. Sequential search. Sequential search

Sequential search. Building Java Programs Chapter 13. Sequential search. Sequential search Sequental search Buldng Java Programs Chapter 13 Searchng and Sortng sequental search: Locates a target value n an array/lst by examnng each element from start to fnsh. How many elements wll t need to

More information

Delay Variation Optimized Traffic Allocation Based on Network Calculus for Multi-path Routing in Wireless Mesh Networks

Delay Variation Optimized Traffic Allocation Based on Network Calculus for Multi-path Routing in Wireless Mesh Networks Appl. Math. Inf. Sc. 7, No. 2L, 467-474 2013) 467 Appled Mathematcs & Informaton Scences An Internatonal Journal http://dx.do.org/10.12785/ams/072l13 Delay Varaton Optmzed Traffc Allocaton Based on Network

More information

High-Boost Mesh Filtering for 3-D Shape Enhancement

High-Boost Mesh Filtering for 3-D Shape Enhancement Hgh-Boost Mesh Flterng for 3-D Shape Enhancement Hrokazu Yagou Λ Alexander Belyaev y Damng We z Λ y z ; ; Shape Modelng Laboratory, Unversty of Azu, Azu-Wakamatsu 965-8580 Japan y Computer Graphcs Group,

More information

Machine Learning: Algorithms and Applications

Machine Learning: Algorithms and Applications 14/05/1 Machne Learnng: Algorthms and Applcatons Florano Zn Free Unversty of Bozen-Bolzano Faculty of Computer Scence Academc Year 011-01 Lecture 10: 14 May 01 Unsupervsed Learnng cont Sldes courtesy of

More information

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

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

More information

Combined Functional Partitioning and Communication Speed Selection for Networked Voltage-Scalable Processors Λ

Combined Functional Partitioning and Communication Speed Selection for Networked Voltage-Scalable Processors Λ Combned Functonal Parttonng and Communcaton Speed Selecton for Networked Voltage-Scalable Processors Λ Jnfeng Lu, Pa H. Chou, Nader Bagherzadeh epartment of Electrcal & Computer Engneerng Unversty of Calforna,

More information

CHAPTER 4 PARALLEL PREFIX ADDER

CHAPTER 4 PARALLEL PREFIX ADDER 93 CHAPTER 4 PARALLEL PREFIX ADDER 4.1 INTRODUCTION VLSI Integer adders fnd applcatons n Arthmetc and Logc Unts (ALUs), mcroprocessors and memory addressng unts. Speed of the adder often decdes the mnmum

More information

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

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

More information

SLAM Summer School 2006 Practical 2: SLAM using Monocular Vision

SLAM Summer School 2006 Practical 2: SLAM using Monocular Vision SLAM Summer School 2006 Practcal 2: SLAM usng Monocular Vson Javer Cvera, Unversty of Zaragoza Andrew J. Davson, Imperal College London J.M.M Montel, Unversty of Zaragoza. josemar@unzar.es, jcvera@unzar.es,

More information

Concurrent Apriori Data Mining Algorithms

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

More information

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

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

More information

A Hybrid Genetic Algorithm for Routing Optimization in IP Networks Utilizing Bandwidth and Delay Metrics

A Hybrid Genetic Algorithm for Routing Optimization in IP Networks Utilizing Bandwidth and Delay Metrics A Hybrd Genetc Algorthm for Routng Optmzaton n IP Networks Utlzng Bandwdth and Delay Metrcs Anton Redl Insttute of Communcaton Networks, Munch Unversty of Technology, Arcsstr. 21, 80290 Munch, Germany

More information

CSE 326: Data Structures Quicksort Comparison Sorting Bound

CSE 326: Data Structures Quicksort Comparison Sorting Bound CSE 326: Data Structures Qucksort Comparson Sortng Bound Bran Curless Sprng 2008 Announcements (5/14/08) Homework due at begnnng of class on Frday. Secton tomorrow: Graded homeworks returned More dscusson

More information

Analysis of Collaborative Distributed Admission Control in x Networks

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

More information

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

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

More information

Research Article. ISSN (Print) s k and. d k rate of k -th flow, source node and

Research Article. ISSN (Print) s k and. d k rate of k -th flow, source node and Scholars Journal of Engneerng and Technology (SJET) Sch. J. Eng. Tech., 2015; 3(4A):343-350 Scholars Academc and Scentfc Publsher (An Internatonal Publsher for Academc and Scentfc Resources) www.saspublsher.com

More information

CSCI 104 Sorting Algorithms. Mark Redekopp David Kempe

CSCI 104 Sorting Algorithms. Mark Redekopp David Kempe CSCI 104 Sortng Algorthms Mark Redekopp Davd Kempe Algorthm Effcency SORTING 2 Sortng If we have an unordered lst, sequental search becomes our only choce If we wll perform a lot of searches t may be benefcal

More information

Virtual Machine Migration based on Trust Measurement of Computer Node

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

More information

Network Delays and Link Capacities in Application-Specific Wormhole NoCs

Network Delays and Link Capacities in Application-Specific Wormhole NoCs Network Delays and Lnk Capactes n Applcaton-Specfc Wormhole NoCs Zvka Guz 1, Isask'har Walter 1, Evgeny Bolotn 1, Israel Cdon 2, Ran Gnosar 2, Avnoam Kolodny 2 Electrcal Engneerng Department, Technon,

More information

A Multilevel Analytical Placement for 3D ICs

A Multilevel Analytical Placement for 3D ICs A Multlevel Analytcal Placement for 3D ICs Jason Cong, and Guoje Luo Computer Scence Department Unversty of Calforna, Los Angeles Calforna NanoSystems Insttute Los Angeles, CA 90095, USA Tel : (30) 06-775

More information

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

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

More information

A SYSTOLIC APPROACH TO LOOP PARTITIONING AND MAPPING INTO FIXED SIZE DISTRIBUTED MEMORY ARCHITECTURES

A SYSTOLIC APPROACH TO LOOP PARTITIONING AND MAPPING INTO FIXED SIZE DISTRIBUTED MEMORY ARCHITECTURES A SYSOLIC APPROACH O LOOP PARIIONING AND MAPPING INO FIXED SIZE DISRIBUED MEMORY ARCHIECURES Ioanns Drosts, Nektaros Kozrs, George Papakonstantnou and Panayots sanakas Natonal echncal Unversty of Athens

More information

Parallel Inverse Halftoning by Look-Up Table (LUT) Partitioning

Parallel Inverse Halftoning by Look-Up Table (LUT) Partitioning Parallel Inverse Halftonng by Look-Up Table (LUT) Parttonng Umar F. Sddq and Sadq M. Sat umar@ccse.kfupm.edu.sa, sadq@kfupm.edu.sa KFUPM Box: Department of Computer Engneerng, Kng Fahd Unversty of Petroleum

More information

On the Network Partitioning of Large Urban Transportation Networks

On the Network Partitioning of Large Urban Transportation Networks On the etwor Parttonng of Large Urban Transportaton etwors Hamdeh Etemadna and Khaled Abdelghany Abstract Ths paper ams at developng a traffc networ parttonng mechansm for dstrbuted traffc management applcatons.

More information

Multi-objective Design Optimization of MCM Placement

Multi-objective Design Optimization of MCM Placement Proceedngs of the 5th WSEAS Int. Conf. on Instrumentaton, Measurement, Crcuts and Systems, Hangzhou, Chna, Aprl 6-8, 26 (pp56-6) Mult-objectve Desgn Optmzaton of MCM Placement Chng-Ma Ko ab, Yu-Jung Huang

More information

A fast algorithm for color image segmentation

A fast algorithm for color image segmentation Unersty of Wollongong Research Onlne Faculty of Informatcs - Papers (Arche) Faculty of Engneerng and Informaton Scences 006 A fast algorthm for color mage segmentaton L. Dong Unersty of Wollongong, lju@uow.edu.au

More information

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

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

More information

An Optimization Algorithm for Minimum Connected Dominating Set Problem in Wireless Sensor Network

An Optimization Algorithm for Minimum Connected Dominating Set Problem in Wireless Sensor Network IEMS Vol. 10, No. 3, pp. 221-231, September 2011. An Optmzaton Algorthm for Mnmum Connected Domnatng Set Problem n Wreless Sensor Network Namsu Ahn Defense Agency for Technology and Qualty 275-18, Boksu-dong,

More information

R s s f. m y s. SPH3UW Unit 7.3 Spherical Concave Mirrors Page 1 of 12. Notes

R s s f. m y s. SPH3UW Unit 7.3 Spherical Concave Mirrors Page 1 of 12. Notes SPH3UW Unt 7.3 Sphercal Concave Mrrors Page 1 of 1 Notes Physcs Tool box Concave Mrror If the reflectng surface takes place on the nner surface of the sphercal shape so that the centre of the mrror bulges

More information

Improvement of Spatial Resolution Using BlockMatching Based Motion Estimation and Frame. Integration

Improvement of Spatial Resolution Using BlockMatching Based Motion Estimation and Frame. Integration Improvement of Spatal Resoluton Usng BlockMatchng Based Moton Estmaton and Frame Integraton Danya Suga and Takayuk Hamamoto Graduate School of Engneerng, Tokyo Unversty of Scence, 6-3-1, Nuku, Katsuska-ku,

More information

CHAPTER 2 PROPOSED IMPROVED PARTICLE SWARM OPTIMIZATION

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

More information

Efficient Load-Balanced IP Routing Scheme Based on Shortest Paths in Hose Model. Eiji Oki May 28, 2009 The University of Electro-Communications

Efficient Load-Balanced IP Routing Scheme Based on Shortest Paths in Hose Model. Eiji Oki May 28, 2009 The University of Electro-Communications Effcent Loa-Balance IP Routng Scheme Base on Shortest Paths n Hose Moel E Ok May 28, 2009 The Unversty of Electro-Communcatons Ok Lab. Semnar, May 28, 2009 1 Outlne Backgroun on IP routng IP routng strategy

More information

Cost-efficient deployment of distributed software services

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

More information

LECTURE : MANIFOLD LEARNING

LECTURE : MANIFOLD LEARNING LECTURE : MANIFOLD LEARNING Rta Osadchy Some sldes are due to L.Saul, V. C. Raykar, N. Verma Topcs PCA MDS IsoMap LLE EgenMaps Done! Dmensonalty Reducton Data representaton Inputs are real-valued vectors

More information

An Euler Path Based Technique for Deadlock-free Multicasting

An Euler Path Based Technique for Deadlock-free Multicasting An Euler Path Based Technque for Deadlockfree Multcastng Ndh Agrawal and C.P.Ravkumar Department of Electrcal Engneerng Indan Insttute of Technology New Delh 006 {ndh,rkumar}@ee.td.ernet.n Abstract The

More information

5 The Primal-Dual Method

5 The Primal-Dual Method 5 The Prmal-Dual Method Orgnally desgned as a method for solvng lnear programs, where t reduces weghted optmzaton problems to smpler combnatoral ones, the prmal-dual method (PDM) has receved much attenton

More information

Classifier Selection Based on Data Complexity Measures *

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

More information

A New Approach For the Ranking of Fuzzy Sets With Different Heights

A New Approach For the Ranking of Fuzzy Sets With Different Heights New pproach For the ankng of Fuzzy Sets Wth Dfferent Heghts Pushpnder Sngh School of Mathematcs Computer pplcatons Thapar Unversty, Patala-7 00 Inda pushpndersnl@gmalcom STCT ankng of fuzzy sets plays

More information

Priority-Based Scheduling Algorithm for Downlink Traffics in IEEE Networks

Priority-Based Scheduling Algorithm for Downlink Traffics in IEEE Networks Prorty-Based Schedulng Algorthm for Downlnk Traffcs n IEEE 80.6 Networks Ja-Mng Lang, Jen-Jee Chen, You-Chun Wang, Yu-Chee Tseng, and Bao-Shuh P. Ln Department of Computer Scence Natonal Chao-Tung Unversty,

More information