Negative Binomial Distribution Based Reputation Direct Update for Wireless Sensor Networks.pdfVIP

JOURNALOFELECTRONICSCIENCEANDTECHNOLOG VOL．9，NO．2，JUNE2011 16l NegativeBinomialDistributionBasedReputation DirectUpdateforW irelessSensorNetworks ZheW ei，Meng-ShuHou，andFangW ang Abstra —_In the reputation modeling ofwireless importanttoolsthathavebeen used inmany fieldslike sensor networks (WSNs) many literatures have economicsand ecommerce．have been introduced into proposed creativereputation indirectupdatemethods， W SNs．In agent-based trustand reputation management such asreputation integration，discounting，aging to (ATRM)，reputationisdefinedashteperceptionthata eliminate， and filtering malicious reputation personhasofanother’sintentionandtrustisviewedas information． However, few have discussed the beliefthatoneentitybelievesthattheotherwillactina reputationdirectupdate．Inthispaper,basedonsound certainway ．Thereputationmetricisconstructedbased statisticaltheories，a negativebinominaldistribution onthestatisticalpropertiesorobservationconsistencyof method in thereputation directupdate forW SNsis sensory data．W henasensornodeproducessensorydata proposed．ResuItsshow thattheproposed method is withstatisticalpropertieshtataredeviatedfrom thenorm． moresuitableandtime-savingofrthereputationupdate itsreputationisconsideredtarnished．accordingly。thisnode oftheresourceconstraintW SNsand canimprovethe becomes lessturswtorthy ．According to thisturstor computationpoweref6leiencyaswel1． reputation mechna ism，notonly the performance and ／ndex TermsBeta distribution，negativebinomial robustnessofthenewtork can be enhanced，butalso a distribution，reputation，wirelesssensornewt orks． stimulating effectofr cooperation amongnodescan be