简介：Wereviewthestateoftheartandourperspectivesonsiliconandhybridsiliconphotonicdevicesforopticalinterconnectsindatacenters.Afterabriefdiscussionofthekeyrequirementsforintra-datacenteropticalinterconnects,weproposeawavelength-division-multiplexing(WDM)-basedopticalinterconnectforintra-datacenterapplications.Followingourproposedinterconnectsconfiguration,thebulkofthereviewemphasizesrecentdevelopmentsconcerningon-chiphybridsiliconmicrolasersandWDMtransmitters,andsiliconphotonicswitchfabricsforintra-datacenters.ForhybridsiliconmicrolasersandWDMtransmitters,weoutlinetheremainingchallengesandkeyissuestowardrealizinglowpowerconsumption,directmodulation,andintegrationofmultiwavelengthmicrolaserarrays.Forsiliconphotonicswitchfabrics,wereviewvarioustopologiesandconfigurationsofhigh-port-countN-by-NswitchfabricsusingMach–Zehnderinterferometersandmicroringresonatorsasswitchelements,anddiscusstheirprospectstowardpracticalimplementationswithactivereconfiguration.Forthemicroring-basedswitchfabrics,wereviewrecentdevelopmentsofactivestabilizationschemesatthesubsystemlevel.Last,weoutlineseverallargechallengesandproblemsforsiliconandhybridsiliconphotonicstomeetforintra-datacenterapplicationsandproposepotentialsolutions.

简介：Withthemulti-tierpricingschemeprovidedbymostofthecloudserviceproviders（CSPs）,theclouduserstypicallyselectahighenoughtransmissionserviceleveltoensurethequalityofservice（QoS）,duetotheseverepenaltyofmissingthetransmissiondeadline.Thisleadstotheso-calledover-provisioningproblem,whichincreasesthetransmissioncostoftheclouduser.Giventhefactthatcloudusersmaynotbeawareoftheirtrafficdemandbeforeaccessingthenetwork,theover-provisioningproblembecomesmoreserious.Inthispaper,weinvestigatehowtoreducethetransmissioncostfromtheperspectiveofcloudusers,especiallywhentheyarenotawareoftheirtrafficdemandbeforethetransmissiondeadline.Thekeyideaistosplitalong-termtransmissionrequestintoseveralshortones.Byselectingthemostsuitabletransmissionservicelevelforeachshort-termrequest,acost-efiqcientinter-datacentertransmissionservicelevelselectionframeworkisobtained.Wefurtherformulatethetransmissionservicelevelselectionproblemasalinearprogrammingproblemandresolveitinanon-linestylewithLyapunovoptimization.Weevaluatetheproposedapproachwithrealtrafficdata.Theexperimentalresultsshowthatourmethodcanreducethetransmissioncostbyupto65.04%.