欢迎来到天天文库
浏览记录
ID:51931880
大小:204.48 KB
页数:6页
时间:2020-03-19
《国际会议演讲稿.doc》由会员上传分享,免费在线阅读,更多相关内容在应用文档-天天文库。
1、自我介绍Thankyou,Mr./Ms.Chair./professorMynameissangqian.Iamveryhonoredtobeheretodooralpresentation.IamaMasterstudentfromHohaiUniversityandIamcurrentlydoingsomeresearchonphysicallayersecurity.Today,Iwouldliketosharewithyousomeofmyresearchonrelayselectionincooperat
2、ivecommunication.(external/ekˈstərnəl;ɪkˈstərnəl/)内容安排:Mypresentationincludesthesefiveparts.First,somebackgroundinformationaboutthisresearch;Second,systemmodelwehavedone;Third,NN-basedrelayselectionschemewehaveproposedForth,SimulationandresultsanalysisAndlast,
3、someconclusionswehavegotP4Partone,introductionFirstly,Iwouldliketogiveyouabitofbackground.Differingfromthetraditionalcryptographictechniquesbasedonsecretkeys,wecanmakeuseofwirelesschannelcharacteristicstoenhancephysicallayersecurity.Cooperativecommunicationhas
4、beenwidelyrecognizedasaneffectivewaytocombatwirelessfadingandprovidediversitygainwhichisoneoftheresearchhotspots.Machinelearningasanemergingtechnologyhasbeenwidelyappliedinimageprocessing,cancerprediction,stockanalysisandotherfields.Sowhynottryitinwirelesscomm
5、unication?P5:Next,IwanttotalkalittlebitaboutpresentstudyRecentstudiesondeeplearningforwirelesscommunicationsystemshaveproposedalternativeapproachestoenhancecertainpartsoftheconventionalcommunicationsystemsuchasmodulationrecognition、channelencodinganddecoding、c
6、hannelestimationanddetectionandanautoencoderwhichcanreplacethetotalsystemwithanovelarchitecture【modulationrecognition:AnNNarchitectureformodulationrecognitionthatconsistsofa4-layerNNandtwotwo-layerNNs。channelencodinganddecoding:AplainDNNarchitectureforchanneld
7、ecodingtodecodekbitsmessagesfromNbitsnoisycodewords。channelestimationanddetection:Adense-Netforsymbol-to-symboldetectioncanadoptlongshort-termmemory(LSTM)todetectanestimatedsymbol.Autoencoder:theautoencodercanrepresenttheentirecommunicationsystemandjointlyopti
8、mizethetransmitterandreceiveroveranAWGNchannel.】P6Sowhydidweconductthisresearch?Well,wewanttoexploitthepotentialbenefitsofdeeplearninginenhancingphysicallayersecurityincooperative(
此文档下载收益归作者所有