Design and Analysis of Algorithms: First Mediterranean Conference on Algorithms, MedAlg 2012, Kibbutz Ein Gedi, Israel, December 3-5, 2012. Proceedings

269 Pages·2012·4.97 MB·English

by  Amihood Amir

Lecture Notes in Computer Science 7659
CommencedPublicationin1973
FoundingandFormerSeriesEditors:
GerhardGoos,JurisHartmanis,andJanvanLeeuwen
EditorialBoard
DavidHutchison
LancasterUniversity,UK
TakeoKanade
CarnegieMellonUniversity,Pittsburgh,PA,USA
JosefKittler
UniversityofSurrey,Guildford,UK
JonM.Kleinberg
CornellUniversity,Ithaca,NY,USA
AlfredKobsa
UniversityofCalifornia,Irvine,CA,USA
FriedemannMattern
ETHZurich,Switzerland
JohnC.Mitchell
StanfordUniversity,CA,USA
MoniNaor
WeizmannInstituteofScience,Rehovot,Israel
OscarNierstrasz
UniversityofBern,Switzerland
C.PanduRangan
IndianInstituteofTechnology,Madras,India
BernhardSteffen
TUDortmundUniversity,Germany
MadhuSudan
MicrosoftResearch,Cambridge,MA,USA
DemetriTerzopoulos
UniversityofCalifornia,LosAngeles,CA,USA
DougTygar
UniversityofCalifornia,Berkeley,CA,USA
GerhardWeikum
MaxPlanckInstituteforInformatics,Saarbruecken,Germany
Guy Even Dror Rawitz (Eds.)
Design and Analysis
of Algorithms
First Mediterranean Conference
on Algorithms, MedAlg 2012
Kibbutz Ein Gedi, Israel, December 3-5, 2012
Proceedings
1 3
VolumeEditors
GuyEven
DrorRawitz
Tel-AvivUniversity
SchoolofElectricalEngineering
Tel-Aviv67789,Israel
E-mail:{guy,rawitz}@eng.tau.ac.il
ISSN0302-9743 e-ISSN1611-3349
ISBN978-3-642-34861-7 e-ISBN978-3-642-34862-4
DOI10.1007/978-3-642-34862-4
SpringerHeidelbergDordrechtLondonNewYork
LibraryofCongressControlNumber:2012951175
CRSubjectClassification(1998):F.2.1-2,G.2.2-3,E.1,G.1.1-2,G.1.6,G.1.10,I.2.8,
I.6.1,I.6.3,G.4
LNCSSublibrary:SL1–TheoreticalComputerScienceandGeneralIssues
©Springer-VerlagBerlinHeidelberg2012
Thisworkissubjecttocopyright.Allrightsarereserved,whetherthewholeorpartofthematerialis
concerned,specificallytherightsoftranslation,reprinting,re-useofillustrations,recitation,broadcasting,
reproductiononmicrofilmsorinanyotherway,andstorageindatabanks.Duplicationofthispublication
orpartsthereofispermittedonlyundertheprovisionsoftheGermanCopyrightLawofSeptember9,1965,
initscurrentversion,andpermissionforusemustalwaysbeobtainedfromSpringer.Violationsareliable
toprosecutionundertheGermanCopyrightLaw.
Theuseofgeneraldescriptivenames,registerednames,trademarks,etc.inthispublicationdoesnotimply,
evenintheabsenceofaspecificstatement,thatsuchnamesareexemptfromtherelevantprotectivelaws
andregulationsandthereforefreeforgeneraluse.
Typesetting:Camera-readybyauthor,dataconversionbyScientificPublishingServices,Chennai,India
Printedonacid-freepaper
SpringerispartofSpringerScience+BusinessMedia(www.springer.com)
Preface
ThisvolumecontainsthepaperspresentedatMedAlg2012:FirstMediterranean
ConferenceonAlgorithmsheldduringDecember3–5,2012,inKibbutzEinGedi.
There were 44 submissions. Each submission was reviewed by at least two,
andonaveragethree,ProgramCommittee members.The committee decidedto
accept 18 papers.
Four invited speakers volunteered to present talks at the conference: Yosi
Azar, Artur Czumaj, Yishay Mansour, and David Peleg.
We would like to thank the ProgramCommittee members and the reviewers
fortheirhelpinselectingthepapersintheseproceedings.Wethankthe authors
for submitting their papers. We thank Moti Medina for his devoted service as
our Publicity Chair. Gabriel Scalosub served as the Local Chair and took close
care of organizing the meeting.
We would like to thank the financial support of I-CORE ALGO: The Israeli
CenterofResearchExcellenceinAlgorithmsheadedbyYishayMansour.Special
thankstoSaritCohenShalevfromI-COREALGOfortheassitanceinorganizing
the conference.
The EasyChair system was used for the submission of the papers, their re-
viewing, and the generation of these proceedings.
September 2012 Guy Even
Dror Rawitz
Organization
Program Committee
Hannah Bast University of Freiburg, Germany
Niv Buchbinder Open University of Israel
Matthias Englert University of Warwick, UK
Guy Even Tel Aviv University, Israel
Pierre Fraigniaud CNRS and University of Paris 7, France
Chien-Chung Huang Humboldt-Universita¨t zu Berlin, Germany
Giuseppe Italiano Rome University “Tor Vergata”, Italy
Ilan Newman Haifa University, Israel
Dror Rawitz Tel Aviv University, Israel
Liam Roditty Bar-Ilan University, Israel
Christian Scheideler University of Paderborn, Germany
Baruch Schieber IBM T.J. Watson Research Center, USA
Tami Tamir The Interdisciplinary Center, Israel
Anke van Zuylen MaxPlanckInstituteforInformatics,Germany
Berthold Vo¨cking RWTH Aachen University, Germany
Peter Widmayer ETH Zurich, Switzerland
Christos Zaroliagis Computer Technology Institute and
University of Patras,Greece
Steering Committee
Pankaj Agarwal Duke University, USA
Imrich Chlamtac University of Trento, Italy
Alberto Marchetti-Spaccamela Sapienza University of Rome, Italy
David Peleg Weizmann Institute, Israel
Michael Segal BGU, Israel
Paul Spirakis University of Patras, Greece
Roger Wattenhofer ETH, Switzerland
Local Arrangements Chair
Gabriel Scalosub Ben-Gurion University of the Negev, Israel
Publicity Chair
Moti Medina Tel Aviv University, Israel
VIII Organization
Additional Reviewers
Abed, Fidaa Marathe, Madhav
Aharoni, Ron Medina, Moti
Antoniadis, Antonios Michail, Panagiotis
Bansal, Nikhil Mihalak, Matus
Brandes, Ulrik Nielsen, Frank
Cacchiani, Valentina Nussbaum, Yahav
Chatzigiannakis, Ioannis Ogierman, Adrian
Ediger, David Ott, Sebastian
Epstein, Leah Patt-Shamir, Boaz
Foschini, Luca Rabinovich, Yuri
Giakkoupis, George Regnier, Mireille
Grigni, Michelangelo Sach, Benjamin
Haeupler, Bernhard Saket, Rishi
Hassidim, Avinatan Schmid, Stefan
Hermelin, Danny Schwartz, Roy
Hruz, Tomas Shende, Sunil
Hu¨llmann, Martina Smorodinsky, Shakhar
Keller, Orgad Solomon, Shay
Kniesburges, Sebastian Tsichlas, Kostas
Kontogiannis,Spyros Viola, Emanuele
Lovett, Shachar Wahlstro¨m, Magnus
Maheshwari, Anil Wong, Prudence W.H.
Table of Contents
Detecting Approximate Periodic Patterns ........................... 1
Amihood Amir, Alberto Apostolico, Estrella Eisenberg,
Gad M. Landau, Avivit Levy, and Noa Lewenstein
Graph Expansion Analysis for Communication Costs of Fast
Rectangular Matrix Multiplication ................................. 13
Grey Ballard, James Demmel, Olga Holtz, Benjamin Lipshitz, and
Oded Schwartz
Multicast Routing for Energy Minimization Using Speed Scaling ....... 37
Nikhil Bansal, Anupam Gupta, Ravishankar Krishnaswamy,
Viswanath Nagarajan, Kirk Pruhs, and Cliff Stein
Reoptimization of the Minimum Total Flow-Time Scheduling
Problem ........................................................ 52
Guy Baram and Tami Tamir
Energy Efficient Caching for Phase-Change Memory.................. 67
Neal Barcelo, Miao Zhou, Daniel Cole, Michael Nugent, and
Kirk Pruhs
Shortest-Elapsed-Time-Firston a Multiprocessor..................... 82
Neal Barcelo, Sungjin Im, Benjamin Moseley, and Kirk Pruhs
Efficient Route Compression for Hybrid Route Planning .............. 93
Gernot Veit Batz, Robert Geisberger, Dennis Luxen,
Peter Sanders, and Roman Zubkov
Multipath Spanners via Fault-Tolerant Spanners ..................... 108
Shiri Chechik, Quentin Godfroy, and David Peleg
Constant Thresholds Can Make Target Set Selection Tractable ........ 120
Morgan Chopin, Andr´e Nichterlein, Rolf Niedermeier, and
Mathias Weller
Faster Variance Computation for Patterns with Gaps................. 134
Fabio Cunial
Enhancing the Computation of Distributed Shortest Paths on Real
Dynamic Networks............................................... 148
Gianlorenzo D’Angelo, Mattia D’Emidio, Daniele Frigioni, and
Daniele Romano
X Table of Contents
Experimental Analysis of Rumor Spreading in Social Networks ........ 159
Benjamin Doerr, Mahmoud Fouz, and Tobias Friedrich
A Randomised Approximation Algorithm for the Partial Vertex Cover
Problem in Hypergraphs .......................................... 174
Mourad El Ouali, Helena Fohlin, and Anand Srivastav
Simulation-BasedAnalysisofTopologyControlAlgorithmsforWireless
Ad Hoc Networks ................................................ 188
Fabian Fuchs, Markus Vo¨lker, and Dorothea Wagner
Cache-Oblivious Dictionaries and Multimaps with Negligible Failure
Probability...................................................... 203
Michael T. Goodrich, Daniel S. Hirschberg,
Michael Mitzenmacher, and Justin Thaler
An Efficient Generator for Clustered Dynamic Random Networks ...... 219
Robert G¨orke, Roland Kluge, Andrea Schumm,
Christian Staudt, and Dorothea Wagner
Slow Down and Sleep for Profit in Online Deadline Scheduling......... 234
Peter Kling, Andreas Cord-Landwehr, and Frederik Mallmann-Trenn
FIFO Queueing Policies for Packets with Heterogeneous Processing..... 248
Kirill Kogan, Alejandro Lo´pez-Ortiz, Sergey I. Nikolenko,
Alexander V. Sirotkin, and Denis Tugaryov
Author Index.................................................. 261
Detecting Approximate Periodic Patterns
Amihood Amir1,2,(cid:2), Alberto Apostolico3,4,(cid:2)(cid:2), Estrella Eisenberg1,
Gad M. Landau5,6,(cid:2)(cid:2)(cid:2), Avivit Levy7,8,†, and Noa Lewenstein9
1 Departmentof Computer Science, Bar-Ilan University,
Ramat-Gan 52900, Israel
[email protected]
2 Departmentof Computer Science, Johns HopkinsUniversity,
Baltimore, MD 21218
3 College of Computing, Georgia Instituteof Technology,
801 AtlanticDrive, Atlanta, GA 30318, USA
[email protected]
4 Dipartimento di Ingegneria dell’ Informazione, Universit`a diPadova,
ViaGradenigo 6/A, 35131 Padova, Italy
5 Department of Computer Science, Universityof Haifa,
Mount Carmel, Haifa 31905, Israel
[email protected]
6 Department of Computer Scienceand Engineering,
Polytechnic Instituteof New York University,
6 Metrotech Center, Brooklyn, NY 11201
7 Department of Software Engineering, ShenkarCollege,
12 AnnaFrank,Ramat-Gan, Israel
[email protected]
8 CRI,Haifa University,Mount Carmel, Haifa 31905, Israel
9 NetanyaCollege, Netanya,Israel
[email protected]
Abstract. Given (cid:2) ∈ [0,1), the (cid:2)-Relative Error Periodic Pattern
Problem (REPP) is thefollowing:
INPUT: Ann-long sequenceS of numberssi ∈N in increasing order.
OUTPUT:Thelongest (cid:2)-relativeerrorperiodicpattern,i.e., thelongest
subsequencesi1,si2,...,sik ofS,forwhichthereexistsanumberpsuch
thattheabsolutedifferencebetweenanytwoconsecutivenumbersinthe
subsequenceis at least p and at most p(1+(cid:2)).
The best known algorithm for this problem has O(n3) time complexity.
This bound is too high for large inputs in practice. In this paper we
give a new algorithm for finding the longest (cid:2)-relative error periodic
pattern(theREPPproblem).Ourmethodisbasedon atransformation
(cid:2) Partly supported by NSFgrant CCR-09-04581, ISFgrant 347/09, and BSF grant
2008217.
(cid:2)(cid:2) Partly supported byBSF grant 2008217.
(cid:2)(cid:2)(cid:2) Partly supported by the National Science Foundation Award 0904246, Israel Sci-
enceFoundationgrant347/09,Yahoo,GrantNo.2008217fromtheUnitedStates-
Israel Binational ScienceFoundation (BSF) and DFG.
† Partly supported byISFgrant 347/09.
G.EvenandD.Rawitz(Eds.):MedAlg2012,LNCS7659,pp.1–12,2012.
(cid:2)c Springer-VerlagBerlinHeidelberg2012
2 A. Amiret al.
oftheinputsequenceintoadifferentrepresentation:the(cid:2)-activemaximal
intervalslist L,definedinthispaper.Weshowthatthetransformationof
StothelistLcanbedoneefficiently(quadraticinnandlinearinthesize
ofL)andprovethatouralgorithmislinearinthesizeofL.Thisenables
us to prove that our algorithm works in sub-cubic time on inputs for
whichthebestknownalgorithm worksinO(n3)time.Moreover,though
it may happenthat ouralgorithm would still becubic, it is neverworse
than the known O(n3)-algorithm and in many situations its complexity
is O(n2) time.
1 Introduction
Many real world phenomena have a particular type of event that repeats pe-
riodically during a certain period of time. Examples of highly periodic events
includeroadtrafficpeaks[5],loadpeaksonwebservers[6],monitoringeventsin
computer networks [1] and many others. Finding periodicity in real-world data
often leads to useful insights by shedding light on the structure of the data,
and giving a basis to predicting future events. Moreover, in some applications
periodic patterns can point out a problem. In a computer network, for exam-
ple, repeating errormessagescanindicate a misconfiguration,orevena security
intrusion such as a port scan [4].
However, such periodic patterns typically occur temporarily, and need not
persistthroughoutthewholeperiodoftimecoveredbytheeventlog.Sinceshort
(approximate) periodic patterns will appear in any sequence of random events
that is long enough, a periodic pattern is more interesting the more repetitions
it contains. Therefore, finding the longest periodic pattern, which contains the
largest number of repetitions of the same event, is of interest.
The input data we are given in such cases can be modeled as a sequence of
events, each associated with a timestamp. Since we study repetition of events
of the same type, we can treat each type of event separately. Thus, the input
consists of a sequence S of n distinct numbers s1,s2,...,sn in increasing order,
which are the times at which an event of a particular type has occurred. A
periodic pattern then corresponds to an (approximate) arithmetic progression
in this sequence. For a given sequence of numbers s1,s2,...,sk, si ∈ S, the
differences si+1−si of consecutive numbers in the sequence are called periods.
In most real-world applications, the timestamps which are given as input are
imprecise. Hence, no exact arithmetic progression may be present in the input,
and it is necessary to allow some slack in the periodic patterns.
Though finding approximate periodic patterns is a widely studied subject
in data mining (e.g. [3, 7–9]), bioinformatics (for example tandem repeats in
genomic sequences), or astronomy (Lomb-Scargle periodograms), the input is
rather different from the one considered in this paper and applying existing
methods in these fields directly or through data conversion is inadequate or
inefficient. In this paper, we follow the recent framework of [2] to the study of
approximate periodic arithmetic progressions.