SlideShare una empresa de Scribd logo
1 de 9
Descargar para leer sin conexión
International Journal of Computer Networks & Communications (IJCNC) Vol.6, No.3, May 2014
DOI : 10.5121/ijcnc.2014.6304 41
A COMPARISON OF EFFICIENT
ALGORITHMS FOR SCHEDULING
PARALLEL DATA REDISTRIBUTION
Marios-Evangelos Kogias and Timotheos Aslanidis
National Technical University of Athens, Athens, Greece
ABSTRACT
Data redistribution in parallel is an often-addressed issue in modern computer networks. In this context, we
study the case of data redistribution over a switching network. Data from the source stations need to be
transferred to the destination stations in the minimum time possible. Unfortunately the time required to
complete the transfer is burdened by each switching and thus producing an optimal schedule is proven to
be computationally intractable. For the purposes of this paper we consider two algorithms, which have
been proved to be very efficient in the past. To get improved results in comparison to previous approaches,
we propose splitting the data in two clusters depending on the size of the data to be transferred. To prove
the efficiency of our approach we ran experiments on all three algorithms, comparing the time span of the
schedules produced as well as the running times to produce those schedules. The test cases we ran
indicate that not only our newly proposed algorithm yields better results in terms of the schedule produced
but runs faster as well.
KEYWORDS
parallel, distributed, package, scheduling, setup cost, approximation, preemption
1. INTRODUCTION
As the need for communication and dissemination of information increases in modern technology
based societies, so does the need for faster and more efficient networks and routing of packages
between stations. In this context the study of parallel machines (parallel supercomputers,
symmetric multiprocessors, multicomputer clusters and IP router switch fabrics), communication
backbones interconnecting the machines and the transfer of large packets of data between them,
has become an issue of major importance. As information loads continue to increase rapidly, it is
expected that the need for well scheduled data transfers that decrease time and resource usage,
will keep on being an often addressed subject for many computer scientists and engineers.
In this manuscript and in the context of redistributing data between parallel machines using a
centralized switch (Figure1), the problem studied is the Preemptive Bipartite Scheduling problem
(encountered as PBS in the literature). Given a set of n1 source stations and a set of n2 destination
stations, we are required to send across data streams, each initiated from a specific source, to
reach an also prespecified target. The duration of each data stream is also predetermined for all
data streams scheduled for transfer. Restrictions in the systems considered, are that no source may
send data towards more than one destination at any time, nor may a destination station receive,
from more than one source station at any time. Messages are sent in packages and to enhance
transfer speed we are allowed to preempt any package and continue transfer of any part of that
package at a later time. Unfortunately, since the system has to reconfigure after each preemption,
International Journal of Computer Networks & Communications (IJCNC) Vol.6, No.3, May 2014
42
any interruption of packages transfer will come with a time cost. Information on the data remains
that were not transferred along with the main part of the package, have to be saved and a new
setup has to be initiated for the next package to start transferring. Consequently prior to sending
any of the packages there will be a setup overhead. We consider this overhead to be constant for
all transfer initiations. In this paper we aim to minimize the duration of the aforementioned
process.
Figure 1. Crossbar switch interconnection network
2. PREVIOUS RESULTS
PBS has been proven to be NP-Hard in [8], and 4/3-ε inapproximable for any ε>0, unless P=NP in
[4].
As PBS algorithms can be implemented in various applications, many polynomial time
algorithms have been designed to produce solutions close to the optimal, found in [1], [4], [5],
[6]. The optimal approximation ratio proven so far is
1
2
d 1
−
+
, where d is the setup cost. Proof of
that can be found in [1]. Tests on the performance of numerous algorithms producing near-
optimal solutions are presented in [4], [5], [6] and [10].
The optimal solution can be found in polynomial time if the setup cost is zero and in the case we
are only interested in minimizing the number of preemptions [8]. Finally, in [1] can be found a
polynomial time algorithm to calculate the optimal solution for a class of graphs that the authors
define.
The performance of our algorithm is compared to that of 2 algorithms found in bibliography:
• A-PBS(d+1), which has the best approximation ratio proven so far [1].
• A1, which appears to perform very well in practice [6].
To schedule each data package, A-PBS(d+1) rounds up the time of each transfer to the closest
multiple of d+1 and calculates the package reducing the workload of each station to the minimum
multiple of d+1.
Whereas, A1 computes an arbitrary package with a maximum number of data transfers and
decides how to preempt by calculating a lower bound to the transfer duration of the remaining
packages plus the cost of the current package, to be the minimum possible.
International Journal of Computer Networks & Communications (IJCNC) Vol.6, No.3, May 2014
43
3. GRAPH REPRESENTATION AND NOTATIONS
Our data representation will be through a bipartite graph G(V,U,E). V will be the set of source
stations, U the set of receiving stations while E, the set of edges, will correspond to the data loads
that have to be transferred from V to U. A weight (or cost) c(v,u), will be assigned to each of the
edges e=(v,u), to denote the time required to transfer the data from node v to node u. Edge
weights are considered to be non-negative integers.
Furthermore the following notation will be used: ∆=∆(G)=max{
v V
max(deg(v))
∈
,
u U
max(deg(u))
∈
},
that is, ∆ will denote the maximum number of data transfers from or to any of the stations.
The function t: V∪U→Z+ will denote the total workload of any station, namely t(v)=
u U
c(v,u)
∈
∑
for any v∈V or t(u)=
v V
c(v,u)
∈
∑ for any u∈U.
W=W(G)=max{
v V
max(t(v))
∈
,
u U
max(t(u))
∈
}, that is W will denote the maximum total transfer time
from or to any station.
d∈Z*
+ will denote the overhead to start the next transfer.
The objective function to be minimized is C(G,d)=
S
i
i 1
T(p )
=
∑ +d·S, where S is the number of data
packages and T(pi) is the time required for each package pi, to be transferred.
A lower bound to the optimal solution is W(G)+d·∆(G). Yet, this lower bound is not always
achievable as shown in [6].
4. THE SPLIT GRAPH ALGORITHM
For the purposes of our algorithm the initial graph is split in two parts. GM comprises edges of
weight at least d and Gm contains all edges of weight less than d. Our main concern for GM is to
keep reducing the workload for each of the stations, achieving the minimum transfer time
possible, whereas in the case of Gm, where edge weights are small in comparison to d, we aim in
minimizing the number of preemptions. The intuition in designing this algorithm is that for data
transfers of long duration, priority on how to schedule has the duration of the data transfer rather
than the number of preemptions, whilst for data transfers of shortest duration prioritized is the
minimization of the number of preemptions. In particular:
The Split-Graph Algorithm (SGA)
Step 1: Split the initial graph G(V,U,E) in two bipartite graphs Gm(Vm,Um,Em) and
GM(VM,UM,EM), where Vm=VM=V, Um=UM=U and Em contains all edges of weight less than d,
EM contains all edges of weight d or more. Clearly in this initiation step E=Em∪EM and
Em∩EM=∅.
Step 2: Use subroutine1 to find a maximal matching M, in GM.
Step 3: Use subroutine 2 to calculate the weight of the matching to be removed. Remove the
corresponding parts of the edges.
Step 4: If possible, add edges to M, from Em to maximize |M| and remove them from Em.
Step 5: Move edges of weight less than d, from the graph induced by step 3 to Em.
Step 6: Repeat steps 2 to 5 until all edges initially in EM have been completely removed.
Step 7: Use subroutine 3 to calculate ∆m maximum matchings in Gm, where ∆m is the degree of
Gm.
International Journal of Computer Networks & Communications (IJCNC) Vol.6, No.3, May 2014
44
Step 8: Schedule the data packets as calculated in steps 2, 3 and 7.
Subroutine 1:
Step 1: M=∅ (Initialization of the matching).
Step 2: For each node w∈VM∪UM calculate t(w).
Step 3: Sort all nodes w∈VM∪UM in decreasing order of t(w). Let L be the induced list of nodes.
Step 4: Let w0 be the 1st node to appear in L. Run sequential search in L to find the 1st neighbour
of w0 appearing in L. Denote that neighbour by w1.
Step 5: M←M∪{w0, w1}.
Step 6: Remove w0, w1 from L.
Step 7: Repeat steps 2, to 6 until M becomes maximal.
Subroutine 2:
Step 1: For each edge e=(v,u) of the matching M, with corresponding weight c(e) calculate what
the value W(Ge) of the induced graph Ge(Ve,Ue,Ee) would be if all edge weights in the matching
were to be reduced by c(e). In this case edges in M, of cost less than c(e) would be completely
removed.
Step2: For all e∈M, set r(e)=c(e)+W(Ge) and denote by e0 the edge such that
r(e0)=min{r(e)|e∈M} and c(e0)=max{c(e)|r(e)=r(e0)}.
Step 3: For each edge e, in M set its new weight c(e)=
0 0c(e) c(e ), if c(e) c(e )
0, otherwise
 − >


.
Subroutine 3:
Step 1: Add nodes and edges to make Gm a regular graph of degree ∆m. New edges will be of zero
weight. In a regular graph, all nodes will be of the same degree.
Step 2: Calculate a maximum matching Mm in Gm and remove all edges of Mm from Gm. Gm’s
degree will now be reduced by 1.
Step 3: Repeat step 2 until Gm=∅.
Example:
Consider matrix A to be the adjacency matrix of G(V, U, E), where V={a1, b1, c1, d1}, U={a2, b2,
c2}. Rows 1, 2, 3 and 4, show the times required to transfer the data from nodes a1, b1, c1, d1
respectively, while columns 1, 2 and 3 stand for the durations needed to transfer the data to a2, b2,
c2, respectively. Suppose d=3.
5 3 2
1 4 3
A
2 5 0
4 2 5
 
 
 =
 
 
 
Let AM be the adjacency matrix corresponding to GM and Am be the adjacency matrix of Gm.
Then:
M
5 3 0
0 4 3
A
0 5 0
4 0 5
 
 
 =
 
 
 
m
0 0 2
1 0 0
A
2 0 0
0 2 0
 
 
 =
 
 
 
International Journal of Computer Networks & Communications (IJCNC) Vol.6, No.3, May 2014
45
t(a1)=8, t(b1)=7, t(c1)=5, t(d1)=9, t(a2)=9, t(b2)=12, t(c2)=8.
sorting the nodes: b2, d1, a2, a1, c2, b1, c1.
M={{b2,a1}, {d1,c2}}, e0={d1,c2}, r(e0)=14.
Add edge {b1,a2}∈Em to M. Schedule a transfer of duration 5
Total time for this iteration including preemption cost is 8
The induced matrices after the 1st
iteration are:
M
5 0 0
0 4 3
A
0 5 0
4 0 0
 
 
 =
 
 
 
, m
0 0 2
0 0 0
A
2 0 0
0 2 0
 
 
 =
 
 
 
t(a1)=5, t(b1)=7, t(c1)=5, t(d1)=4, t(a2)=9, t(b2)=9, t(c2)=3
sorting the nodes: a2, b2, b1, a1, c1, d1, c2
M={{a2,a1},{b2,b1}} e0={a2,a1}, r(e0)=10.
No additional edge from Em. Schedule a transfer of duration 5
Total time for this iteration including preemption cost is 8
The induced matrices after the 2nd
iteration are:
M
0 0 0
0 0 3
A
0 5 0
4 0 0
 
 
 =
 
 
 
, m
0 0 2
0 0 0
A
2 0 0
0 2 0
 
 
 =
 
 
 
t(a1)=0, t(b1)=3, t(c1)=5, t(d1)=4, t(a2)=4, t(b2)=5, t(c2)=3
sorting the nodes: c1, b2, d1, a2, b1, c2, a1.
M={{c1,b2},{d1,a2},{b1,c2}} e0={c1,b2}, r(e0)=5.
No additional edge from Em. Schedule a transfer of duration 5
Total time for this iteration including preemption cost is 8.
All edges in EM are now removed, proceeding to step 7 of SGA:
m
0 0 2
0 0 0
A
2 0 0
0 2 0
 
 
 =
 
 
 
The corresponding maximum matching is M={{a1,c2}, {c1,a2}, {d1,b2}}
The duration of the scheduled transfer is 2.
Total time for this iteration including preemption cost is 5.
Total time to transfer the data is 29, while the lower bound is 26.
5. COMPARING THE SCHEDULES
One thousand test cases have been ran for a 15 source-15 destination system for values of setup
cost varying from 1 to 100 and message durations varying from 0 to 50. Figure 2 represents each
algorithm’s performance in terms of proximity to the lower bound. SGA performs significantly
International Journal of Computer Networks & Communications (IJCNC) Vol.6, No.3, May 2014
46
better than both A1 and A-PBS(d+1) and as the overhead increases it shows an increasingly
improved performance. It is important to mention that in practice, as information loads
exponentially increase, the number of stations and communication tasks increases and so does the
setup cost. That is in fact the most encountered situation nowadays.
Figure 3 presents the worst performance that each algorithm had depending on the setup cost, in
terms of proximity to the lower bound again. SGA is found to be once again a lot more efficient.
Furthermore, SGA in most cases has a worst case really close to the average showing that its
performance does not fluctuate much, making it an all cases reliable tool for this type of
scheduling.
It is important to stress out that the ratio calculated, is in comparison to the lower bound and not
to the optimal value of the objective function. Therefore the schedules produced are in fact a lot
closer to the optimal and the ratios a lot closer to 1.
Figure 2. Average (cost/lower bound) comparison
International Journal of Computer Networks & Communications (IJCNC) Vol.6, No.3, May 2014
47
Figure 3. Worst (cost/lower bound) comparison
6. COMPARING THE RUNNING TIMES
To compare time complexities for the 3 algorithms we will assume that |E|=m and we will denote
max{|V|,|U|}=n.
All three algorithms compared, will need O(m) iterations to produce a schedule as it is certain that
each iteration will result in the complete removal of at least one edge.
For each iteration the most time consuming subroutine of A-PBS(d+1) is finding a maximum
matching which will need O( n m) [12]. Therefore the overall complexity of A-PBS(d+1) is
O( n m2
)
As far as A1 is concerned, determining how to preempt is the most time consuming subroutine for
each iteration. For each of the edge weights in the matching, it calculates for each node v, deg(v)
and t(v). Each of these three processes requires O(n), therefore the complexity of each iteration is
O(n3
) making the overall time complexity O(n3
m).
Finally, the complexity of SGA is determined by the maximum matching process that takes place
in step2 of subroutine3 and consequently the overall complexity is the same as that of A-
PBS(d+1), namely O( n m2
).
As |E| varies from O(n) to O(n2
), it is clear that A-PBS(d+1) and SGA work faster than A1. Yet,
SGA will only be using the maximum matching algorithm after most of the graphs edges are
removed. Consequently in practice it runs somewhat faster.
International Journal of Computer Networks & Communications (IJCNC) Vol.6, No.3, May 2014
48
7. CONCLUSIONS AND FUTURE WORK
Our newly presented algorithm (SGA), has proven to produce more efficient schedules. We
believe that the idea of splitting the initial graph in parts can be further researched and depending
on the magnitude of the edges as well as the setup cost, the Split-Graph Algorithm’s efficiency
can be further improved. An approximation ratio for SGA could be established to be less than 2.
Exploiting to a greater extend algorithms that provide optimal solutions for special instances of
the problem might also yield interesting new approximation algorithms. Finally, lifting limitations
of the problem or introducing new ones could help in identifying new classes of graphs for which
polynomial algorithms might provide an optimal schedule.
REFERENCES
[1] F. Afrati, T. Aslanidis, E. Bampis, I. Milis, Scheduling in Switching Networks with Set-up
Delays. Journal of Combinatorial Optimization, vol. 9, issue 1, p.49-57, Feb 2005.
[2] P. B. Bhat, V. K. Prasanna and C. S. Raghavendra, Block Cyclic Redistribution over
Heterogeneous Networks, in 11th
International Conference on Parallel and Distributed Computing
Systems (PDCS 1998), 1998.
[3] G. Bongiovanni, D. Coppersmith and C. K.Wong, An optimal time slot assignment for an
SS/TDMA system with variable number of transponders, IEEE Trans. Commun. vol. 29, p. 721-
726, 1981.
[4] J. Cohen, E. Jeannot, N. Padoy and F. Wagner, Messages Scheduling for Parallel Data
Redistribution between Clusters, IEEE Transactions on Parallel and Distributed Systems, vol. 17,
Number 10, p. 1163, 2006.
[5] J. Cohen, E. Jeannot, N. Padoy, Parallel Data Redistribution Over a Backbone, Technical Report
RR-4725, INRIA-Lorraine, February 2003.
[6] P. Crescenzi, X. Deng, C. H. Papadimitriou, On approximating a scheduling problem, Journal of
Combinatorial Optimization, vol. 5, p. 287-297, 2001.
[7] I. S. Gopal, G. Bongiovanni, M. A. Bonucelli, D. T. Tang, C. K. Wong, An optimal switching
algorithm for multibeam satellite systems with variable bandwidth beams, IEEE Trans. Commun.
vol. 30, p. 2475-2481, Nov. 1982.
[8] I. S. Gopal, C. K. Wong, Minimizing the number of switchings in an SS/TDMA system IEEE
Trans. Commun. vol. 33, p. 497-501, 1985.
[9] T. Inukai, An efficient SS/TDMA time slot assignment algorithm IEEE Trans. Commun. vol 27, p.
1449-1455, Oct. 1979.
[10] E. Jeannot and F. Wagner, Two fast and efficient message scheduling algorithms for data
redistribution over a backbone, 18th International Parallel and Distributed Processing
Symposium, 2004.
[11] A. Kesselman and K. Kogan, Nonpreemptive Scheduling of Optical Switches, IEEE Transactions
in Communications, vol. 55, number 6, p. 1212, 2007.
[12] S. Micali and V. V. Vazirani, An O( | V | |E|) algorithm for finding a maximum matching in
general graphs, in Proc. 21st
Ann IEEE Symp. Foundations of Computer Science, p. 17-
27, 1980.
[13] K. S. Natarajan and S. B. Calo, Time slot assignment in an SS/TDMA system with minimum
switchings IBM Res. Rep. 1981.
[14] B. Towles and W. J. Dally, Guaranteed Scheduling of Switches with Configuration Overhead, in
Proc. Twenty-First Annual Joint Conference of the IEEE Computer and Communications Societies
INFOCOM ’02. pp. 342-351, June 2002.
International Journal of Computer Networks & Communications (IJCNC) Vol.6, No.3, May 2014
49
Authors
Timotheos Aslanidis was born in Athens, Greece in 1974. He received his Mathematics
degree from the University of Athens in 1997 and a master's degree in computer science
in 2001. He is currently a phd candidate at the National and Technical University of
Athens in the department of electrical and computer engineering. His research interests
comprise but are not limited to computer theory, number theory, network algorithms and
data mining algorithms.
Marios-Evangelos Kogias was born in Athens in 1991. Currently, he is an undergraduate
student at the National Technical University of Athens pursuing a diploma in Electrical
and Computer Engineering. He is a passionate coder and interested mainly in operating
systems, computer networks and algorithms.

Más contenido relacionado

La actualidad más candente

timeSeriesClassificationLDA
timeSeriesClassificationLDAtimeSeriesClassificationLDA
timeSeriesClassificationLDAKellen Betts
 
Image encryption technique incorporating wavelet transform and hash integrity
Image encryption technique incorporating wavelet transform and hash integrityImage encryption technique incorporating wavelet transform and hash integrity
Image encryption technique incorporating wavelet transform and hash integrityeSAT Journals
 
IJCER (www.ijceronline.com) International Journal of computational Engineerin...
IJCER (www.ijceronline.com) International Journal of computational Engineerin...IJCER (www.ijceronline.com) International Journal of computational Engineerin...
IJCER (www.ijceronline.com) International Journal of computational Engineerin...ijceronline
 
Optimization of Collective Communication in MPICH
Optimization of Collective Communication in MPICH Optimization of Collective Communication in MPICH
Optimization of Collective Communication in MPICH Lino Possamai
 
STATE SPACE GENERATION FRAMEWORK BASED ON BINARY DECISION DIAGRAM FOR DISTRIB...
STATE SPACE GENERATION FRAMEWORK BASED ON BINARY DECISION DIAGRAM FOR DISTRIB...STATE SPACE GENERATION FRAMEWORK BASED ON BINARY DECISION DIAGRAM FOR DISTRIB...
STATE SPACE GENERATION FRAMEWORK BASED ON BINARY DECISION DIAGRAM FOR DISTRIB...csandit
 
ICC paper
ICC paperICC paper
ICC paperQi Chen
 

La actualidad más candente (16)

Chap9 slides
Chap9 slidesChap9 slides
Chap9 slides
 
Chap4 slides
Chap4 slidesChap4 slides
Chap4 slides
 
D0341015020
D0341015020D0341015020
D0341015020
 
timeSeriesClassificationLDA
timeSeriesClassificationLDAtimeSeriesClassificationLDA
timeSeriesClassificationLDA
 
Image encryption technique incorporating wavelet transform and hash integrity
Image encryption technique incorporating wavelet transform and hash integrityImage encryption technique incorporating wavelet transform and hash integrity
Image encryption technique incorporating wavelet transform and hash integrity
 
post119s1-file3
post119s1-file3post119s1-file3
post119s1-file3
 
IJCER (www.ijceronline.com) International Journal of computational Engineerin...
IJCER (www.ijceronline.com) International Journal of computational Engineerin...IJCER (www.ijceronline.com) International Journal of computational Engineerin...
IJCER (www.ijceronline.com) International Journal of computational Engineerin...
 
Optimization of Collective Communication in MPICH
Optimization of Collective Communication in MPICH Optimization of Collective Communication in MPICH
Optimization of Collective Communication in MPICH
 
Chap8 slides
Chap8 slidesChap8 slides
Chap8 slides
 
robio-2014-falquez
robio-2014-falquezrobio-2014-falquez
robio-2014-falquez
 
Aggarwal Draft
Aggarwal DraftAggarwal Draft
Aggarwal Draft
 
Chap4 slides
Chap4 slidesChap4 slides
Chap4 slides
 
STATE SPACE GENERATION FRAMEWORK BASED ON BINARY DECISION DIAGRAM FOR DISTRIB...
STATE SPACE GENERATION FRAMEWORK BASED ON BINARY DECISION DIAGRAM FOR DISTRIB...STATE SPACE GENERATION FRAMEWORK BASED ON BINARY DECISION DIAGRAM FOR DISTRIB...
STATE SPACE GENERATION FRAMEWORK BASED ON BINARY DECISION DIAGRAM FOR DISTRIB...
 
ICC paper
ICC paperICC paper
ICC paper
 
Chap10 slides
Chap10 slidesChap10 slides
Chap10 slides
 
End sem
End semEnd sem
End sem
 

Similar a A comparison of efficient algorithms for scheduling parallel data redistribution

IMPROVING SCHEDULING OF DATA TRANSMISSION IN TDMA SYSTEMS
IMPROVING SCHEDULING OF DATA TRANSMISSION IN TDMA SYSTEMSIMPROVING SCHEDULING OF DATA TRANSMISSION IN TDMA SYSTEMS
IMPROVING SCHEDULING OF DATA TRANSMISSION IN TDMA SYSTEMScscpconf
 
An Open Shop Approach in Approximating Optimal Data Transmission Duration in ...
An Open Shop Approach in Approximating Optimal Data Transmission Duration in ...An Open Shop Approach in Approximating Optimal Data Transmission Duration in ...
An Open Shop Approach in Approximating Optimal Data Transmission Duration in ...cscpconf
 
MESSAGE ROUTING IN WIRELESS AND MOBILE NETWORKS USING TDMA TECHNOLOGY
MESSAGE ROUTING IN WIRELESS AND MOBILE NETWORKS USING TDMA TECHNOLOGYMESSAGE ROUTING IN WIRELESS AND MOBILE NETWORKS USING TDMA TECHNOLOGY
MESSAGE ROUTING IN WIRELESS AND MOBILE NETWORKS USING TDMA TECHNOLOGYijwmn
 
MESSAGE ROUTING IN WIRELESS AND MOBILE NETWORKS USING TDMA TECHNOLOGY
MESSAGE ROUTING IN WIRELESS AND MOBILE NETWORKS USING TDMA TECHNOLOGYMESSAGE ROUTING IN WIRELESS AND MOBILE NETWORKS USING TDMA TECHNOLOGY
MESSAGE ROUTING IN WIRELESS AND MOBILE NETWORKS USING TDMA TECHNOLOGYijwmn
 
Dce a novel delay correlation
Dce a novel delay correlationDce a novel delay correlation
Dce a novel delay correlationijdpsjournal
 
A CRITICAL IMPROVEMENT ON OPEN SHOP SCHEDULING ALGORITHM FOR ROUTING IN INTER...
A CRITICAL IMPROVEMENT ON OPEN SHOP SCHEDULING ALGORITHM FOR ROUTING IN INTER...A CRITICAL IMPROVEMENT ON OPEN SHOP SCHEDULING ALGORITHM FOR ROUTING IN INTER...
A CRITICAL IMPROVEMENT ON OPEN SHOP SCHEDULING ALGORITHM FOR ROUTING IN INTER...IJCNCJournal
 
DCE: A NOVEL DELAY CORRELATION MEASUREMENT FOR TOMOGRAPHY WITH PASSIVE REAL...
DCE: A NOVEL DELAY CORRELATION  MEASUREMENT FOR TOMOGRAPHY WITH PASSIVE  REAL...DCE: A NOVEL DELAY CORRELATION  MEASUREMENT FOR TOMOGRAPHY WITH PASSIVE  REAL...
DCE: A NOVEL DELAY CORRELATION MEASUREMENT FOR TOMOGRAPHY WITH PASSIVE REAL...ijdpsjournal
 
An improved spfa algorithm for single source shortest path problem using forw...
An improved spfa algorithm for single source shortest path problem using forw...An improved spfa algorithm for single source shortest path problem using forw...
An improved spfa algorithm for single source shortest path problem using forw...IJMIT JOURNAL
 
International Journal of Managing Information Technology (IJMIT)
International Journal of Managing Information Technology (IJMIT)International Journal of Managing Information Technology (IJMIT)
International Journal of Managing Information Technology (IJMIT)IJMIT JOURNAL
 
An improved spfa algorithm for single source shortest path problem using forw...
An improved spfa algorithm for single source shortest path problem using forw...An improved spfa algorithm for single source shortest path problem using forw...
An improved spfa algorithm for single source shortest path problem using forw...IJMIT JOURNAL
 
An Ant Algorithm for Solving QoS Multicast Routing Problem
An Ant Algorithm for Solving QoS Multicast Routing ProblemAn Ant Algorithm for Solving QoS Multicast Routing Problem
An Ant Algorithm for Solving QoS Multicast Routing ProblemCSCJournals
 
Algorithm Finding Maximum Concurrent Multicommodity Linear Flow with Limited ...
Algorithm Finding Maximum Concurrent Multicommodity Linear Flow with Limited ...Algorithm Finding Maximum Concurrent Multicommodity Linear Flow with Limited ...
Algorithm Finding Maximum Concurrent Multicommodity Linear Flow with Limited ...IJCNCJournal
 
Scheduling Using Multi Objective Genetic Algorithm
Scheduling Using Multi Objective Genetic AlgorithmScheduling Using Multi Objective Genetic Algorithm
Scheduling Using Multi Objective Genetic Algorithmiosrjce
 
cis97003
cis97003cis97003
cis97003perfj
 
Dynamic Economic Dispatch Assessment Using Particle Swarm Optimization Technique
Dynamic Economic Dispatch Assessment Using Particle Swarm Optimization TechniqueDynamic Economic Dispatch Assessment Using Particle Swarm Optimization Technique
Dynamic Economic Dispatch Assessment Using Particle Swarm Optimization TechniquejournalBEEI
 
OPTIMIZED TASK ALLOCATION IN SENSOR NETWORKS
OPTIMIZED TASK ALLOCATION IN SENSOR NETWORKSOPTIMIZED TASK ALLOCATION IN SENSOR NETWORKS
OPTIMIZED TASK ALLOCATION IN SENSOR NETWORKSZac Darcy
 
SLIDING WINDOW SUM ALGORITHMS FOR DEEP NEURAL NETWORKS
SLIDING WINDOW SUM ALGORITHMS FOR DEEP NEURAL NETWORKSSLIDING WINDOW SUM ALGORITHMS FOR DEEP NEURAL NETWORKS
SLIDING WINDOW SUM ALGORITHMS FOR DEEP NEURAL NETWORKSIJCI JOURNAL
 
Parallel implementation of geodesic distance transform with application in su...
Parallel implementation of geodesic distance transform with application in su...Parallel implementation of geodesic distance transform with application in su...
Parallel implementation of geodesic distance transform with application in su...Tuan Q. Pham
 
Drobics, m. 2001: datamining using synergiesbetween self-organising maps and...
Drobics, m. 2001:  datamining using synergiesbetween self-organising maps and...Drobics, m. 2001:  datamining using synergiesbetween self-organising maps and...
Drobics, m. 2001: datamining using synergiesbetween self-organising maps and...ArchiLab 7
 

Similar a A comparison of efficient algorithms for scheduling parallel data redistribution (20)

IMPROVING SCHEDULING OF DATA TRANSMISSION IN TDMA SYSTEMS
IMPROVING SCHEDULING OF DATA TRANSMISSION IN TDMA SYSTEMSIMPROVING SCHEDULING OF DATA TRANSMISSION IN TDMA SYSTEMS
IMPROVING SCHEDULING OF DATA TRANSMISSION IN TDMA SYSTEMS
 
An Open Shop Approach in Approximating Optimal Data Transmission Duration in ...
An Open Shop Approach in Approximating Optimal Data Transmission Duration in ...An Open Shop Approach in Approximating Optimal Data Transmission Duration in ...
An Open Shop Approach in Approximating Optimal Data Transmission Duration in ...
 
MESSAGE ROUTING IN WIRELESS AND MOBILE NETWORKS USING TDMA TECHNOLOGY
MESSAGE ROUTING IN WIRELESS AND MOBILE NETWORKS USING TDMA TECHNOLOGYMESSAGE ROUTING IN WIRELESS AND MOBILE NETWORKS USING TDMA TECHNOLOGY
MESSAGE ROUTING IN WIRELESS AND MOBILE NETWORKS USING TDMA TECHNOLOGY
 
MESSAGE ROUTING IN WIRELESS AND MOBILE NETWORKS USING TDMA TECHNOLOGY
MESSAGE ROUTING IN WIRELESS AND MOBILE NETWORKS USING TDMA TECHNOLOGYMESSAGE ROUTING IN WIRELESS AND MOBILE NETWORKS USING TDMA TECHNOLOGY
MESSAGE ROUTING IN WIRELESS AND MOBILE NETWORKS USING TDMA TECHNOLOGY
 
Dce a novel delay correlation
Dce a novel delay correlationDce a novel delay correlation
Dce a novel delay correlation
 
A CRITICAL IMPROVEMENT ON OPEN SHOP SCHEDULING ALGORITHM FOR ROUTING IN INTER...
A CRITICAL IMPROVEMENT ON OPEN SHOP SCHEDULING ALGORITHM FOR ROUTING IN INTER...A CRITICAL IMPROVEMENT ON OPEN SHOP SCHEDULING ALGORITHM FOR ROUTING IN INTER...
A CRITICAL IMPROVEMENT ON OPEN SHOP SCHEDULING ALGORITHM FOR ROUTING IN INTER...
 
DCE: A NOVEL DELAY CORRELATION MEASUREMENT FOR TOMOGRAPHY WITH PASSIVE REAL...
DCE: A NOVEL DELAY CORRELATION  MEASUREMENT FOR TOMOGRAPHY WITH PASSIVE  REAL...DCE: A NOVEL DELAY CORRELATION  MEASUREMENT FOR TOMOGRAPHY WITH PASSIVE  REAL...
DCE: A NOVEL DELAY CORRELATION MEASUREMENT FOR TOMOGRAPHY WITH PASSIVE REAL...
 
An improved spfa algorithm for single source shortest path problem using forw...
An improved spfa algorithm for single source shortest path problem using forw...An improved spfa algorithm for single source shortest path problem using forw...
An improved spfa algorithm for single source shortest path problem using forw...
 
International Journal of Managing Information Technology (IJMIT)
International Journal of Managing Information Technology (IJMIT)International Journal of Managing Information Technology (IJMIT)
International Journal of Managing Information Technology (IJMIT)
 
An improved spfa algorithm for single source shortest path problem using forw...
An improved spfa algorithm for single source shortest path problem using forw...An improved spfa algorithm for single source shortest path problem using forw...
An improved spfa algorithm for single source shortest path problem using forw...
 
An Ant Algorithm for Solving QoS Multicast Routing Problem
An Ant Algorithm for Solving QoS Multicast Routing ProblemAn Ant Algorithm for Solving QoS Multicast Routing Problem
An Ant Algorithm for Solving QoS Multicast Routing Problem
 
Algorithm Finding Maximum Concurrent Multicommodity Linear Flow with Limited ...
Algorithm Finding Maximum Concurrent Multicommodity Linear Flow with Limited ...Algorithm Finding Maximum Concurrent Multicommodity Linear Flow with Limited ...
Algorithm Finding Maximum Concurrent Multicommodity Linear Flow with Limited ...
 
M017327378
M017327378M017327378
M017327378
 
Scheduling Using Multi Objective Genetic Algorithm
Scheduling Using Multi Objective Genetic AlgorithmScheduling Using Multi Objective Genetic Algorithm
Scheduling Using Multi Objective Genetic Algorithm
 
cis97003
cis97003cis97003
cis97003
 
Dynamic Economic Dispatch Assessment Using Particle Swarm Optimization Technique
Dynamic Economic Dispatch Assessment Using Particle Swarm Optimization TechniqueDynamic Economic Dispatch Assessment Using Particle Swarm Optimization Technique
Dynamic Economic Dispatch Assessment Using Particle Swarm Optimization Technique
 
OPTIMIZED TASK ALLOCATION IN SENSOR NETWORKS
OPTIMIZED TASK ALLOCATION IN SENSOR NETWORKSOPTIMIZED TASK ALLOCATION IN SENSOR NETWORKS
OPTIMIZED TASK ALLOCATION IN SENSOR NETWORKS
 
SLIDING WINDOW SUM ALGORITHMS FOR DEEP NEURAL NETWORKS
SLIDING WINDOW SUM ALGORITHMS FOR DEEP NEURAL NETWORKSSLIDING WINDOW SUM ALGORITHMS FOR DEEP NEURAL NETWORKS
SLIDING WINDOW SUM ALGORITHMS FOR DEEP NEURAL NETWORKS
 
Parallel implementation of geodesic distance transform with application in su...
Parallel implementation of geodesic distance transform with application in su...Parallel implementation of geodesic distance transform with application in su...
Parallel implementation of geodesic distance transform with application in su...
 
Drobics, m. 2001: datamining using synergiesbetween self-organising maps and...
Drobics, m. 2001:  datamining using synergiesbetween self-organising maps and...Drobics, m. 2001:  datamining using synergiesbetween self-organising maps and...
Drobics, m. 2001: datamining using synergiesbetween self-organising maps and...
 

Más de IJCNCJournal

Advanced Privacy Scheme to Improve Road Safety in Smart Transportation Systems
Advanced Privacy Scheme to Improve Road Safety in Smart Transportation SystemsAdvanced Privacy Scheme to Improve Road Safety in Smart Transportation Systems
Advanced Privacy Scheme to Improve Road Safety in Smart Transportation SystemsIJCNCJournal
 
April 2024 - Top 10 Read Articles in Computer Networks & Communications
April 2024 - Top 10 Read Articles in Computer Networks & CommunicationsApril 2024 - Top 10 Read Articles in Computer Networks & Communications
April 2024 - Top 10 Read Articles in Computer Networks & CommunicationsIJCNCJournal
 
DEF: Deep Ensemble Neural Network Classifier for Android Malware Detection
DEF: Deep Ensemble Neural Network Classifier for Android Malware DetectionDEF: Deep Ensemble Neural Network Classifier for Android Malware Detection
DEF: Deep Ensemble Neural Network Classifier for Android Malware DetectionIJCNCJournal
 
High Performance NMF Based Intrusion Detection System for Big Data IOT Traffic
High Performance NMF Based Intrusion Detection System for Big Data IOT TrafficHigh Performance NMF Based Intrusion Detection System for Big Data IOT Traffic
High Performance NMF Based Intrusion Detection System for Big Data IOT TrafficIJCNCJournal
 
A Novel Medium Access Control Strategy for Heterogeneous Traffic in Wireless ...
A Novel Medium Access Control Strategy for Heterogeneous Traffic in Wireless ...A Novel Medium Access Control Strategy for Heterogeneous Traffic in Wireless ...
A Novel Medium Access Control Strategy for Heterogeneous Traffic in Wireless ...IJCNCJournal
 
A Topology Control Algorithm Taking into Account Energy and Quality of Transm...
A Topology Control Algorithm Taking into Account Energy and Quality of Transm...A Topology Control Algorithm Taking into Account Energy and Quality of Transm...
A Topology Control Algorithm Taking into Account Energy and Quality of Transm...IJCNCJournal
 
Multi-Server user Authentication Scheme for Privacy Preservation with Fuzzy C...
Multi-Server user Authentication Scheme for Privacy Preservation with Fuzzy C...Multi-Server user Authentication Scheme for Privacy Preservation with Fuzzy C...
Multi-Server user Authentication Scheme for Privacy Preservation with Fuzzy C...IJCNCJournal
 
Advanced Privacy Scheme to Improve Road Safety in Smart Transportation Systems
Advanced Privacy Scheme to Improve Road Safety in Smart Transportation SystemsAdvanced Privacy Scheme to Improve Road Safety in Smart Transportation Systems
Advanced Privacy Scheme to Improve Road Safety in Smart Transportation SystemsIJCNCJournal
 
DEF: Deep Ensemble Neural Network Classifier for Android Malware Detection
DEF: Deep Ensemble Neural Network Classifier for Android Malware DetectionDEF: Deep Ensemble Neural Network Classifier for Android Malware Detection
DEF: Deep Ensemble Neural Network Classifier for Android Malware DetectionIJCNCJournal
 
High Performance NMF based Intrusion Detection System for Big Data IoT Traffic
High Performance NMF based Intrusion Detection System for Big Data IoT TrafficHigh Performance NMF based Intrusion Detection System for Big Data IoT Traffic
High Performance NMF based Intrusion Detection System for Big Data IoT TrafficIJCNCJournal
 
IoT Guardian: A Novel Feature Discovery and Cooperative Game Theory Empowered...
IoT Guardian: A Novel Feature Discovery and Cooperative Game Theory Empowered...IoT Guardian: A Novel Feature Discovery and Cooperative Game Theory Empowered...
IoT Guardian: A Novel Feature Discovery and Cooperative Game Theory Empowered...IJCNCJournal
 
Enhancing Traffic Routing Inside a Network through IoT Technology & Network C...
Enhancing Traffic Routing Inside a Network through IoT Technology & Network C...Enhancing Traffic Routing Inside a Network through IoT Technology & Network C...
Enhancing Traffic Routing Inside a Network through IoT Technology & Network C...IJCNCJournal
 
IoT Guardian: A Novel Feature Discovery and Cooperative Game Theory Empowered...
IoT Guardian: A Novel Feature Discovery and Cooperative Game Theory Empowered...IoT Guardian: A Novel Feature Discovery and Cooperative Game Theory Empowered...
IoT Guardian: A Novel Feature Discovery and Cooperative Game Theory Empowered...IJCNCJournal
 
** Connect, Collaborate, And Innovate: IJCNC - Where Networking Futures Take ...
** Connect, Collaborate, And Innovate: IJCNC - Where Networking Futures Take ...** Connect, Collaborate, And Innovate: IJCNC - Where Networking Futures Take ...
** Connect, Collaborate, And Innovate: IJCNC - Where Networking Futures Take ...IJCNCJournal
 
Enhancing Traffic Routing Inside a Network through IoT Technology & Network C...
Enhancing Traffic Routing Inside a Network through IoT Technology & Network C...Enhancing Traffic Routing Inside a Network through IoT Technology & Network C...
Enhancing Traffic Routing Inside a Network through IoT Technology & Network C...IJCNCJournal
 
Multipoint Relay Path for Efficient Topology Maintenance Algorithm in Optimiz...
Multipoint Relay Path for Efficient Topology Maintenance Algorithm in Optimiz...Multipoint Relay Path for Efficient Topology Maintenance Algorithm in Optimiz...
Multipoint Relay Path for Efficient Topology Maintenance Algorithm in Optimiz...IJCNCJournal
 
March 2024 - Top 10 Read Articles in Computer Networks & Communications
March 2024 - Top 10 Read Articles in Computer Networks & CommunicationsMarch 2024 - Top 10 Read Articles in Computer Networks & Communications
March 2024 - Top 10 Read Articles in Computer Networks & CommunicationsIJCNCJournal
 
Adaptive Multi-Criteria-Based Load Balancing Technique for Resource Allocatio...
Adaptive Multi-Criteria-Based Load Balancing Technique for Resource Allocatio...Adaptive Multi-Criteria-Based Load Balancing Technique for Resource Allocatio...
Adaptive Multi-Criteria-Based Load Balancing Technique for Resource Allocatio...IJCNCJournal
 
Comparative Study of Orchestration using gRPC API and REST API in Server Crea...
Comparative Study of Orchestration using gRPC API and REST API in Server Crea...Comparative Study of Orchestration using gRPC API and REST API in Server Crea...
Comparative Study of Orchestration using gRPC API and REST API in Server Crea...IJCNCJournal
 
Sensing Time Improvement Using Two Stage Detectors for Cognitive Radio System
Sensing Time Improvement Using Two Stage Detectors for Cognitive Radio SystemSensing Time Improvement Using Two Stage Detectors for Cognitive Radio System
Sensing Time Improvement Using Two Stage Detectors for Cognitive Radio SystemIJCNCJournal
 

Más de IJCNCJournal (20)

Advanced Privacy Scheme to Improve Road Safety in Smart Transportation Systems
Advanced Privacy Scheme to Improve Road Safety in Smart Transportation SystemsAdvanced Privacy Scheme to Improve Road Safety in Smart Transportation Systems
Advanced Privacy Scheme to Improve Road Safety in Smart Transportation Systems
 
April 2024 - Top 10 Read Articles in Computer Networks & Communications
April 2024 - Top 10 Read Articles in Computer Networks & CommunicationsApril 2024 - Top 10 Read Articles in Computer Networks & Communications
April 2024 - Top 10 Read Articles in Computer Networks & Communications
 
DEF: Deep Ensemble Neural Network Classifier for Android Malware Detection
DEF: Deep Ensemble Neural Network Classifier for Android Malware DetectionDEF: Deep Ensemble Neural Network Classifier for Android Malware Detection
DEF: Deep Ensemble Neural Network Classifier for Android Malware Detection
 
High Performance NMF Based Intrusion Detection System for Big Data IOT Traffic
High Performance NMF Based Intrusion Detection System for Big Data IOT TrafficHigh Performance NMF Based Intrusion Detection System for Big Data IOT Traffic
High Performance NMF Based Intrusion Detection System for Big Data IOT Traffic
 
A Novel Medium Access Control Strategy for Heterogeneous Traffic in Wireless ...
A Novel Medium Access Control Strategy for Heterogeneous Traffic in Wireless ...A Novel Medium Access Control Strategy for Heterogeneous Traffic in Wireless ...
A Novel Medium Access Control Strategy for Heterogeneous Traffic in Wireless ...
 
A Topology Control Algorithm Taking into Account Energy and Quality of Transm...
A Topology Control Algorithm Taking into Account Energy and Quality of Transm...A Topology Control Algorithm Taking into Account Energy and Quality of Transm...
A Topology Control Algorithm Taking into Account Energy and Quality of Transm...
 
Multi-Server user Authentication Scheme for Privacy Preservation with Fuzzy C...
Multi-Server user Authentication Scheme for Privacy Preservation with Fuzzy C...Multi-Server user Authentication Scheme for Privacy Preservation with Fuzzy C...
Multi-Server user Authentication Scheme for Privacy Preservation with Fuzzy C...
 
Advanced Privacy Scheme to Improve Road Safety in Smart Transportation Systems
Advanced Privacy Scheme to Improve Road Safety in Smart Transportation SystemsAdvanced Privacy Scheme to Improve Road Safety in Smart Transportation Systems
Advanced Privacy Scheme to Improve Road Safety in Smart Transportation Systems
 
DEF: Deep Ensemble Neural Network Classifier for Android Malware Detection
DEF: Deep Ensemble Neural Network Classifier for Android Malware DetectionDEF: Deep Ensemble Neural Network Classifier for Android Malware Detection
DEF: Deep Ensemble Neural Network Classifier for Android Malware Detection
 
High Performance NMF based Intrusion Detection System for Big Data IoT Traffic
High Performance NMF based Intrusion Detection System for Big Data IoT TrafficHigh Performance NMF based Intrusion Detection System for Big Data IoT Traffic
High Performance NMF based Intrusion Detection System for Big Data IoT Traffic
 
IoT Guardian: A Novel Feature Discovery and Cooperative Game Theory Empowered...
IoT Guardian: A Novel Feature Discovery and Cooperative Game Theory Empowered...IoT Guardian: A Novel Feature Discovery and Cooperative Game Theory Empowered...
IoT Guardian: A Novel Feature Discovery and Cooperative Game Theory Empowered...
 
Enhancing Traffic Routing Inside a Network through IoT Technology & Network C...
Enhancing Traffic Routing Inside a Network through IoT Technology & Network C...Enhancing Traffic Routing Inside a Network through IoT Technology & Network C...
Enhancing Traffic Routing Inside a Network through IoT Technology & Network C...
 
IoT Guardian: A Novel Feature Discovery and Cooperative Game Theory Empowered...
IoT Guardian: A Novel Feature Discovery and Cooperative Game Theory Empowered...IoT Guardian: A Novel Feature Discovery and Cooperative Game Theory Empowered...
IoT Guardian: A Novel Feature Discovery and Cooperative Game Theory Empowered...
 
** Connect, Collaborate, And Innovate: IJCNC - Where Networking Futures Take ...
** Connect, Collaborate, And Innovate: IJCNC - Where Networking Futures Take ...** Connect, Collaborate, And Innovate: IJCNC - Where Networking Futures Take ...
** Connect, Collaborate, And Innovate: IJCNC - Where Networking Futures Take ...
 
Enhancing Traffic Routing Inside a Network through IoT Technology & Network C...
Enhancing Traffic Routing Inside a Network through IoT Technology & Network C...Enhancing Traffic Routing Inside a Network through IoT Technology & Network C...
Enhancing Traffic Routing Inside a Network through IoT Technology & Network C...
 
Multipoint Relay Path for Efficient Topology Maintenance Algorithm in Optimiz...
Multipoint Relay Path for Efficient Topology Maintenance Algorithm in Optimiz...Multipoint Relay Path for Efficient Topology Maintenance Algorithm in Optimiz...
Multipoint Relay Path for Efficient Topology Maintenance Algorithm in Optimiz...
 
March 2024 - Top 10 Read Articles in Computer Networks & Communications
March 2024 - Top 10 Read Articles in Computer Networks & CommunicationsMarch 2024 - Top 10 Read Articles in Computer Networks & Communications
March 2024 - Top 10 Read Articles in Computer Networks & Communications
 
Adaptive Multi-Criteria-Based Load Balancing Technique for Resource Allocatio...
Adaptive Multi-Criteria-Based Load Balancing Technique for Resource Allocatio...Adaptive Multi-Criteria-Based Load Balancing Technique for Resource Allocatio...
Adaptive Multi-Criteria-Based Load Balancing Technique for Resource Allocatio...
 
Comparative Study of Orchestration using gRPC API and REST API in Server Crea...
Comparative Study of Orchestration using gRPC API and REST API in Server Crea...Comparative Study of Orchestration using gRPC API and REST API in Server Crea...
Comparative Study of Orchestration using gRPC API and REST API in Server Crea...
 
Sensing Time Improvement Using Two Stage Detectors for Cognitive Radio System
Sensing Time Improvement Using Two Stage Detectors for Cognitive Radio SystemSensing Time Improvement Using Two Stage Detectors for Cognitive Radio System
Sensing Time Improvement Using Two Stage Detectors for Cognitive Radio System
 

Último

Automating Google Workspace (GWS) & more with Apps Script
Automating Google Workspace (GWS) & more with Apps ScriptAutomating Google Workspace (GWS) & more with Apps Script
Automating Google Workspace (GWS) & more with Apps Scriptwesley chun
 
[2024]Digital Global Overview Report 2024 Meltwater.pdf
[2024]Digital Global Overview Report 2024 Meltwater.pdf[2024]Digital Global Overview Report 2024 Meltwater.pdf
[2024]Digital Global Overview Report 2024 Meltwater.pdfhans926745
 
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptxHampshireHUG
 
Partners Life - Insurer Innovation Award 2024
Partners Life - Insurer Innovation Award 2024Partners Life - Insurer Innovation Award 2024
Partners Life - Insurer Innovation Award 2024The Digital Insurer
 
Top 5 Benefits OF Using Muvi Live Paywall For Live Streams
Top 5 Benefits OF Using Muvi Live Paywall For Live StreamsTop 5 Benefits OF Using Muvi Live Paywall For Live Streams
Top 5 Benefits OF Using Muvi Live Paywall For Live StreamsRoshan Dwivedi
 
Unblocking The Main Thread Solving ANRs and Frozen Frames
Unblocking The Main Thread Solving ANRs and Frozen FramesUnblocking The Main Thread Solving ANRs and Frozen Frames
Unblocking The Main Thread Solving ANRs and Frozen FramesSinan KOZAK
 
Salesforce Community Group Quito, Salesforce 101
Salesforce Community Group Quito, Salesforce 101Salesforce Community Group Quito, Salesforce 101
Salesforce Community Group Quito, Salesforce 101Paola De la Torre
 
The Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptxThe Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptxMalak Abu Hammad
 
How to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected WorkerHow to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected WorkerThousandEyes
 
Injustice - Developers Among Us (SciFiDevCon 2024)
Injustice - Developers Among Us (SciFiDevCon 2024)Injustice - Developers Among Us (SciFiDevCon 2024)
Injustice - Developers Among Us (SciFiDevCon 2024)Allon Mureinik
 
GenCyber Cyber Security Day Presentation
GenCyber Cyber Security Day PresentationGenCyber Cyber Security Day Presentation
GenCyber Cyber Security Day PresentationMichael W. Hawkins
 
EIS-Webinar-Prompt-Knowledge-Eng-2024-04-08.pptx
EIS-Webinar-Prompt-Knowledge-Eng-2024-04-08.pptxEIS-Webinar-Prompt-Knowledge-Eng-2024-04-08.pptx
EIS-Webinar-Prompt-Knowledge-Eng-2024-04-08.pptxEarley Information Science
 
Slack Application Development 101 Slides
Slack Application Development 101 SlidesSlack Application Development 101 Slides
Slack Application Development 101 Slidespraypatel2
 
08448380779 Call Girls In Civil Lines Women Seeking Men
08448380779 Call Girls In Civil Lines Women Seeking Men08448380779 Call Girls In Civil Lines Women Seeking Men
08448380779 Call Girls In Civil Lines Women Seeking MenDelhi Call girls
 
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure serviceWhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure servicePooja Nehwal
 
Breaking the Kubernetes Kill Chain: Host Path Mount
Breaking the Kubernetes Kill Chain: Host Path MountBreaking the Kubernetes Kill Chain: Host Path Mount
Breaking the Kubernetes Kill Chain: Host Path MountPuma Security, LLC
 
IAC 2024 - IA Fast Track to Search Focused AI Solutions
IAC 2024 - IA Fast Track to Search Focused AI SolutionsIAC 2024 - IA Fast Track to Search Focused AI Solutions
IAC 2024 - IA Fast Track to Search Focused AI SolutionsEnterprise Knowledge
 
Handwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed textsHandwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed textsMaria Levchenko
 
Finology Group – Insurtech Innovation Award 2024
Finology Group – Insurtech Innovation Award 2024Finology Group – Insurtech Innovation Award 2024
Finology Group – Insurtech Innovation Award 2024The Digital Insurer
 
Exploring the Future Potential of AI-Enabled Smartphone Processors
Exploring the Future Potential of AI-Enabled Smartphone ProcessorsExploring the Future Potential of AI-Enabled Smartphone Processors
Exploring the Future Potential of AI-Enabled Smartphone Processorsdebabhi2
 

Último (20)

Automating Google Workspace (GWS) & more with Apps Script
Automating Google Workspace (GWS) & more with Apps ScriptAutomating Google Workspace (GWS) & more with Apps Script
Automating Google Workspace (GWS) & more with Apps Script
 
[2024]Digital Global Overview Report 2024 Meltwater.pdf
[2024]Digital Global Overview Report 2024 Meltwater.pdf[2024]Digital Global Overview Report 2024 Meltwater.pdf
[2024]Digital Global Overview Report 2024 Meltwater.pdf
 
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
 
Partners Life - Insurer Innovation Award 2024
Partners Life - Insurer Innovation Award 2024Partners Life - Insurer Innovation Award 2024
Partners Life - Insurer Innovation Award 2024
 
Top 5 Benefits OF Using Muvi Live Paywall For Live Streams
Top 5 Benefits OF Using Muvi Live Paywall For Live StreamsTop 5 Benefits OF Using Muvi Live Paywall For Live Streams
Top 5 Benefits OF Using Muvi Live Paywall For Live Streams
 
Unblocking The Main Thread Solving ANRs and Frozen Frames
Unblocking The Main Thread Solving ANRs and Frozen FramesUnblocking The Main Thread Solving ANRs and Frozen Frames
Unblocking The Main Thread Solving ANRs and Frozen Frames
 
Salesforce Community Group Quito, Salesforce 101
Salesforce Community Group Quito, Salesforce 101Salesforce Community Group Quito, Salesforce 101
Salesforce Community Group Quito, Salesforce 101
 
The Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptxThe Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptx
 
How to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected WorkerHow to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected Worker
 
Injustice - Developers Among Us (SciFiDevCon 2024)
Injustice - Developers Among Us (SciFiDevCon 2024)Injustice - Developers Among Us (SciFiDevCon 2024)
Injustice - Developers Among Us (SciFiDevCon 2024)
 
GenCyber Cyber Security Day Presentation
GenCyber Cyber Security Day PresentationGenCyber Cyber Security Day Presentation
GenCyber Cyber Security Day Presentation
 
EIS-Webinar-Prompt-Knowledge-Eng-2024-04-08.pptx
EIS-Webinar-Prompt-Knowledge-Eng-2024-04-08.pptxEIS-Webinar-Prompt-Knowledge-Eng-2024-04-08.pptx
EIS-Webinar-Prompt-Knowledge-Eng-2024-04-08.pptx
 
Slack Application Development 101 Slides
Slack Application Development 101 SlidesSlack Application Development 101 Slides
Slack Application Development 101 Slides
 
08448380779 Call Girls In Civil Lines Women Seeking Men
08448380779 Call Girls In Civil Lines Women Seeking Men08448380779 Call Girls In Civil Lines Women Seeking Men
08448380779 Call Girls In Civil Lines Women Seeking Men
 
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure serviceWhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
 
Breaking the Kubernetes Kill Chain: Host Path Mount
Breaking the Kubernetes Kill Chain: Host Path MountBreaking the Kubernetes Kill Chain: Host Path Mount
Breaking the Kubernetes Kill Chain: Host Path Mount
 
IAC 2024 - IA Fast Track to Search Focused AI Solutions
IAC 2024 - IA Fast Track to Search Focused AI SolutionsIAC 2024 - IA Fast Track to Search Focused AI Solutions
IAC 2024 - IA Fast Track to Search Focused AI Solutions
 
Handwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed textsHandwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed texts
 
Finology Group – Insurtech Innovation Award 2024
Finology Group – Insurtech Innovation Award 2024Finology Group – Insurtech Innovation Award 2024
Finology Group – Insurtech Innovation Award 2024
 
Exploring the Future Potential of AI-Enabled Smartphone Processors
Exploring the Future Potential of AI-Enabled Smartphone ProcessorsExploring the Future Potential of AI-Enabled Smartphone Processors
Exploring the Future Potential of AI-Enabled Smartphone Processors
 

A comparison of efficient algorithms for scheduling parallel data redistribution

  • 1. International Journal of Computer Networks & Communications (IJCNC) Vol.6, No.3, May 2014 DOI : 10.5121/ijcnc.2014.6304 41 A COMPARISON OF EFFICIENT ALGORITHMS FOR SCHEDULING PARALLEL DATA REDISTRIBUTION Marios-Evangelos Kogias and Timotheos Aslanidis National Technical University of Athens, Athens, Greece ABSTRACT Data redistribution in parallel is an often-addressed issue in modern computer networks. In this context, we study the case of data redistribution over a switching network. Data from the source stations need to be transferred to the destination stations in the minimum time possible. Unfortunately the time required to complete the transfer is burdened by each switching and thus producing an optimal schedule is proven to be computationally intractable. For the purposes of this paper we consider two algorithms, which have been proved to be very efficient in the past. To get improved results in comparison to previous approaches, we propose splitting the data in two clusters depending on the size of the data to be transferred. To prove the efficiency of our approach we ran experiments on all three algorithms, comparing the time span of the schedules produced as well as the running times to produce those schedules. The test cases we ran indicate that not only our newly proposed algorithm yields better results in terms of the schedule produced but runs faster as well. KEYWORDS parallel, distributed, package, scheduling, setup cost, approximation, preemption 1. INTRODUCTION As the need for communication and dissemination of information increases in modern technology based societies, so does the need for faster and more efficient networks and routing of packages between stations. In this context the study of parallel machines (parallel supercomputers, symmetric multiprocessors, multicomputer clusters and IP router switch fabrics), communication backbones interconnecting the machines and the transfer of large packets of data between them, has become an issue of major importance. As information loads continue to increase rapidly, it is expected that the need for well scheduled data transfers that decrease time and resource usage, will keep on being an often addressed subject for many computer scientists and engineers. In this manuscript and in the context of redistributing data between parallel machines using a centralized switch (Figure1), the problem studied is the Preemptive Bipartite Scheduling problem (encountered as PBS in the literature). Given a set of n1 source stations and a set of n2 destination stations, we are required to send across data streams, each initiated from a specific source, to reach an also prespecified target. The duration of each data stream is also predetermined for all data streams scheduled for transfer. Restrictions in the systems considered, are that no source may send data towards more than one destination at any time, nor may a destination station receive, from more than one source station at any time. Messages are sent in packages and to enhance transfer speed we are allowed to preempt any package and continue transfer of any part of that package at a later time. Unfortunately, since the system has to reconfigure after each preemption,
  • 2. International Journal of Computer Networks & Communications (IJCNC) Vol.6, No.3, May 2014 42 any interruption of packages transfer will come with a time cost. Information on the data remains that were not transferred along with the main part of the package, have to be saved and a new setup has to be initiated for the next package to start transferring. Consequently prior to sending any of the packages there will be a setup overhead. We consider this overhead to be constant for all transfer initiations. In this paper we aim to minimize the duration of the aforementioned process. Figure 1. Crossbar switch interconnection network 2. PREVIOUS RESULTS PBS has been proven to be NP-Hard in [8], and 4/3-ε inapproximable for any ε>0, unless P=NP in [4]. As PBS algorithms can be implemented in various applications, many polynomial time algorithms have been designed to produce solutions close to the optimal, found in [1], [4], [5], [6]. The optimal approximation ratio proven so far is 1 2 d 1 − + , where d is the setup cost. Proof of that can be found in [1]. Tests on the performance of numerous algorithms producing near- optimal solutions are presented in [4], [5], [6] and [10]. The optimal solution can be found in polynomial time if the setup cost is zero and in the case we are only interested in minimizing the number of preemptions [8]. Finally, in [1] can be found a polynomial time algorithm to calculate the optimal solution for a class of graphs that the authors define. The performance of our algorithm is compared to that of 2 algorithms found in bibliography: • A-PBS(d+1), which has the best approximation ratio proven so far [1]. • A1, which appears to perform very well in practice [6]. To schedule each data package, A-PBS(d+1) rounds up the time of each transfer to the closest multiple of d+1 and calculates the package reducing the workload of each station to the minimum multiple of d+1. Whereas, A1 computes an arbitrary package with a maximum number of data transfers and decides how to preempt by calculating a lower bound to the transfer duration of the remaining packages plus the cost of the current package, to be the minimum possible.
  • 3. International Journal of Computer Networks & Communications (IJCNC) Vol.6, No.3, May 2014 43 3. GRAPH REPRESENTATION AND NOTATIONS Our data representation will be through a bipartite graph G(V,U,E). V will be the set of source stations, U the set of receiving stations while E, the set of edges, will correspond to the data loads that have to be transferred from V to U. A weight (or cost) c(v,u), will be assigned to each of the edges e=(v,u), to denote the time required to transfer the data from node v to node u. Edge weights are considered to be non-negative integers. Furthermore the following notation will be used: ∆=∆(G)=max{ v V max(deg(v)) ∈ , u U max(deg(u)) ∈ }, that is, ∆ will denote the maximum number of data transfers from or to any of the stations. The function t: V∪U→Z+ will denote the total workload of any station, namely t(v)= u U c(v,u) ∈ ∑ for any v∈V or t(u)= v V c(v,u) ∈ ∑ for any u∈U. W=W(G)=max{ v V max(t(v)) ∈ , u U max(t(u)) ∈ }, that is W will denote the maximum total transfer time from or to any station. d∈Z* + will denote the overhead to start the next transfer. The objective function to be minimized is C(G,d)= S i i 1 T(p ) = ∑ +d·S, where S is the number of data packages and T(pi) is the time required for each package pi, to be transferred. A lower bound to the optimal solution is W(G)+d·∆(G). Yet, this lower bound is not always achievable as shown in [6]. 4. THE SPLIT GRAPH ALGORITHM For the purposes of our algorithm the initial graph is split in two parts. GM comprises edges of weight at least d and Gm contains all edges of weight less than d. Our main concern for GM is to keep reducing the workload for each of the stations, achieving the minimum transfer time possible, whereas in the case of Gm, where edge weights are small in comparison to d, we aim in minimizing the number of preemptions. The intuition in designing this algorithm is that for data transfers of long duration, priority on how to schedule has the duration of the data transfer rather than the number of preemptions, whilst for data transfers of shortest duration prioritized is the minimization of the number of preemptions. In particular: The Split-Graph Algorithm (SGA) Step 1: Split the initial graph G(V,U,E) in two bipartite graphs Gm(Vm,Um,Em) and GM(VM,UM,EM), where Vm=VM=V, Um=UM=U and Em contains all edges of weight less than d, EM contains all edges of weight d or more. Clearly in this initiation step E=Em∪EM and Em∩EM=∅. Step 2: Use subroutine1 to find a maximal matching M, in GM. Step 3: Use subroutine 2 to calculate the weight of the matching to be removed. Remove the corresponding parts of the edges. Step 4: If possible, add edges to M, from Em to maximize |M| and remove them from Em. Step 5: Move edges of weight less than d, from the graph induced by step 3 to Em. Step 6: Repeat steps 2 to 5 until all edges initially in EM have been completely removed. Step 7: Use subroutine 3 to calculate ∆m maximum matchings in Gm, where ∆m is the degree of Gm.
  • 4. International Journal of Computer Networks & Communications (IJCNC) Vol.6, No.3, May 2014 44 Step 8: Schedule the data packets as calculated in steps 2, 3 and 7. Subroutine 1: Step 1: M=∅ (Initialization of the matching). Step 2: For each node w∈VM∪UM calculate t(w). Step 3: Sort all nodes w∈VM∪UM in decreasing order of t(w). Let L be the induced list of nodes. Step 4: Let w0 be the 1st node to appear in L. Run sequential search in L to find the 1st neighbour of w0 appearing in L. Denote that neighbour by w1. Step 5: M←M∪{w0, w1}. Step 6: Remove w0, w1 from L. Step 7: Repeat steps 2, to 6 until M becomes maximal. Subroutine 2: Step 1: For each edge e=(v,u) of the matching M, with corresponding weight c(e) calculate what the value W(Ge) of the induced graph Ge(Ve,Ue,Ee) would be if all edge weights in the matching were to be reduced by c(e). In this case edges in M, of cost less than c(e) would be completely removed. Step2: For all e∈M, set r(e)=c(e)+W(Ge) and denote by e0 the edge such that r(e0)=min{r(e)|e∈M} and c(e0)=max{c(e)|r(e)=r(e0)}. Step 3: For each edge e, in M set its new weight c(e)= 0 0c(e) c(e ), if c(e) c(e ) 0, otherwise  − >   . Subroutine 3: Step 1: Add nodes and edges to make Gm a regular graph of degree ∆m. New edges will be of zero weight. In a regular graph, all nodes will be of the same degree. Step 2: Calculate a maximum matching Mm in Gm and remove all edges of Mm from Gm. Gm’s degree will now be reduced by 1. Step 3: Repeat step 2 until Gm=∅. Example: Consider matrix A to be the adjacency matrix of G(V, U, E), where V={a1, b1, c1, d1}, U={a2, b2, c2}. Rows 1, 2, 3 and 4, show the times required to transfer the data from nodes a1, b1, c1, d1 respectively, while columns 1, 2 and 3 stand for the durations needed to transfer the data to a2, b2, c2, respectively. Suppose d=3. 5 3 2 1 4 3 A 2 5 0 4 2 5      =       Let AM be the adjacency matrix corresponding to GM and Am be the adjacency matrix of Gm. Then: M 5 3 0 0 4 3 A 0 5 0 4 0 5      =       m 0 0 2 1 0 0 A 2 0 0 0 2 0      =      
  • 5. International Journal of Computer Networks & Communications (IJCNC) Vol.6, No.3, May 2014 45 t(a1)=8, t(b1)=7, t(c1)=5, t(d1)=9, t(a2)=9, t(b2)=12, t(c2)=8. sorting the nodes: b2, d1, a2, a1, c2, b1, c1. M={{b2,a1}, {d1,c2}}, e0={d1,c2}, r(e0)=14. Add edge {b1,a2}∈Em to M. Schedule a transfer of duration 5 Total time for this iteration including preemption cost is 8 The induced matrices after the 1st iteration are: M 5 0 0 0 4 3 A 0 5 0 4 0 0      =       , m 0 0 2 0 0 0 A 2 0 0 0 2 0      =       t(a1)=5, t(b1)=7, t(c1)=5, t(d1)=4, t(a2)=9, t(b2)=9, t(c2)=3 sorting the nodes: a2, b2, b1, a1, c1, d1, c2 M={{a2,a1},{b2,b1}} e0={a2,a1}, r(e0)=10. No additional edge from Em. Schedule a transfer of duration 5 Total time for this iteration including preemption cost is 8 The induced matrices after the 2nd iteration are: M 0 0 0 0 0 3 A 0 5 0 4 0 0      =       , m 0 0 2 0 0 0 A 2 0 0 0 2 0      =       t(a1)=0, t(b1)=3, t(c1)=5, t(d1)=4, t(a2)=4, t(b2)=5, t(c2)=3 sorting the nodes: c1, b2, d1, a2, b1, c2, a1. M={{c1,b2},{d1,a2},{b1,c2}} e0={c1,b2}, r(e0)=5. No additional edge from Em. Schedule a transfer of duration 5 Total time for this iteration including preemption cost is 8. All edges in EM are now removed, proceeding to step 7 of SGA: m 0 0 2 0 0 0 A 2 0 0 0 2 0      =       The corresponding maximum matching is M={{a1,c2}, {c1,a2}, {d1,b2}} The duration of the scheduled transfer is 2. Total time for this iteration including preemption cost is 5. Total time to transfer the data is 29, while the lower bound is 26. 5. COMPARING THE SCHEDULES One thousand test cases have been ran for a 15 source-15 destination system for values of setup cost varying from 1 to 100 and message durations varying from 0 to 50. Figure 2 represents each algorithm’s performance in terms of proximity to the lower bound. SGA performs significantly
  • 6. International Journal of Computer Networks & Communications (IJCNC) Vol.6, No.3, May 2014 46 better than both A1 and A-PBS(d+1) and as the overhead increases it shows an increasingly improved performance. It is important to mention that in practice, as information loads exponentially increase, the number of stations and communication tasks increases and so does the setup cost. That is in fact the most encountered situation nowadays. Figure 3 presents the worst performance that each algorithm had depending on the setup cost, in terms of proximity to the lower bound again. SGA is found to be once again a lot more efficient. Furthermore, SGA in most cases has a worst case really close to the average showing that its performance does not fluctuate much, making it an all cases reliable tool for this type of scheduling. It is important to stress out that the ratio calculated, is in comparison to the lower bound and not to the optimal value of the objective function. Therefore the schedules produced are in fact a lot closer to the optimal and the ratios a lot closer to 1. Figure 2. Average (cost/lower bound) comparison
  • 7. International Journal of Computer Networks & Communications (IJCNC) Vol.6, No.3, May 2014 47 Figure 3. Worst (cost/lower bound) comparison 6. COMPARING THE RUNNING TIMES To compare time complexities for the 3 algorithms we will assume that |E|=m and we will denote max{|V|,|U|}=n. All three algorithms compared, will need O(m) iterations to produce a schedule as it is certain that each iteration will result in the complete removal of at least one edge. For each iteration the most time consuming subroutine of A-PBS(d+1) is finding a maximum matching which will need O( n m) [12]. Therefore the overall complexity of A-PBS(d+1) is O( n m2 ) As far as A1 is concerned, determining how to preempt is the most time consuming subroutine for each iteration. For each of the edge weights in the matching, it calculates for each node v, deg(v) and t(v). Each of these three processes requires O(n), therefore the complexity of each iteration is O(n3 ) making the overall time complexity O(n3 m). Finally, the complexity of SGA is determined by the maximum matching process that takes place in step2 of subroutine3 and consequently the overall complexity is the same as that of A- PBS(d+1), namely O( n m2 ). As |E| varies from O(n) to O(n2 ), it is clear that A-PBS(d+1) and SGA work faster than A1. Yet, SGA will only be using the maximum matching algorithm after most of the graphs edges are removed. Consequently in practice it runs somewhat faster.
  • 8. International Journal of Computer Networks & Communications (IJCNC) Vol.6, No.3, May 2014 48 7. CONCLUSIONS AND FUTURE WORK Our newly presented algorithm (SGA), has proven to produce more efficient schedules. We believe that the idea of splitting the initial graph in parts can be further researched and depending on the magnitude of the edges as well as the setup cost, the Split-Graph Algorithm’s efficiency can be further improved. An approximation ratio for SGA could be established to be less than 2. Exploiting to a greater extend algorithms that provide optimal solutions for special instances of the problem might also yield interesting new approximation algorithms. Finally, lifting limitations of the problem or introducing new ones could help in identifying new classes of graphs for which polynomial algorithms might provide an optimal schedule. REFERENCES [1] F. Afrati, T. Aslanidis, E. Bampis, I. Milis, Scheduling in Switching Networks with Set-up Delays. Journal of Combinatorial Optimization, vol. 9, issue 1, p.49-57, Feb 2005. [2] P. B. Bhat, V. K. Prasanna and C. S. Raghavendra, Block Cyclic Redistribution over Heterogeneous Networks, in 11th International Conference on Parallel and Distributed Computing Systems (PDCS 1998), 1998. [3] G. Bongiovanni, D. Coppersmith and C. K.Wong, An optimal time slot assignment for an SS/TDMA system with variable number of transponders, IEEE Trans. Commun. vol. 29, p. 721- 726, 1981. [4] J. Cohen, E. Jeannot, N. Padoy and F. Wagner, Messages Scheduling for Parallel Data Redistribution between Clusters, IEEE Transactions on Parallel and Distributed Systems, vol. 17, Number 10, p. 1163, 2006. [5] J. Cohen, E. Jeannot, N. Padoy, Parallel Data Redistribution Over a Backbone, Technical Report RR-4725, INRIA-Lorraine, February 2003. [6] P. Crescenzi, X. Deng, C. H. Papadimitriou, On approximating a scheduling problem, Journal of Combinatorial Optimization, vol. 5, p. 287-297, 2001. [7] I. S. Gopal, G. Bongiovanni, M. A. Bonucelli, D. T. Tang, C. K. Wong, An optimal switching algorithm for multibeam satellite systems with variable bandwidth beams, IEEE Trans. Commun. vol. 30, p. 2475-2481, Nov. 1982. [8] I. S. Gopal, C. K. Wong, Minimizing the number of switchings in an SS/TDMA system IEEE Trans. Commun. vol. 33, p. 497-501, 1985. [9] T. Inukai, An efficient SS/TDMA time slot assignment algorithm IEEE Trans. Commun. vol 27, p. 1449-1455, Oct. 1979. [10] E. Jeannot and F. Wagner, Two fast and efficient message scheduling algorithms for data redistribution over a backbone, 18th International Parallel and Distributed Processing Symposium, 2004. [11] A. Kesselman and K. Kogan, Nonpreemptive Scheduling of Optical Switches, IEEE Transactions in Communications, vol. 55, number 6, p. 1212, 2007. [12] S. Micali and V. V. Vazirani, An O( | V | |E|) algorithm for finding a maximum matching in general graphs, in Proc. 21st Ann IEEE Symp. Foundations of Computer Science, p. 17- 27, 1980. [13] K. S. Natarajan and S. B. Calo, Time slot assignment in an SS/TDMA system with minimum switchings IBM Res. Rep. 1981. [14] B. Towles and W. J. Dally, Guaranteed Scheduling of Switches with Configuration Overhead, in Proc. Twenty-First Annual Joint Conference of the IEEE Computer and Communications Societies INFOCOM ’02. pp. 342-351, June 2002.
  • 9. International Journal of Computer Networks & Communications (IJCNC) Vol.6, No.3, May 2014 49 Authors Timotheos Aslanidis was born in Athens, Greece in 1974. He received his Mathematics degree from the University of Athens in 1997 and a master's degree in computer science in 2001. He is currently a phd candidate at the National and Technical University of Athens in the department of electrical and computer engineering. His research interests comprise but are not limited to computer theory, number theory, network algorithms and data mining algorithms. Marios-Evangelos Kogias was born in Athens in 1991. Currently, he is an undergraduate student at the National Technical University of Athens pursuing a diploma in Electrical and Computer Engineering. He is a passionate coder and interested mainly in operating systems, computer networks and algorithms.