基於深度学习之後5G行动网路—子计划四:基於深度学习之後5G行动边缘雾运算研究
项目来源
台(略)府(略)金(略)B(略)
项目主持人
张(略)
项目受资助机构
台(略)东(略)讯(略)系
财政年度
2(略),(略)9
立项时间
未(略)
项目编号
M(略)1(略)2(略)-(略)3(略)1(略)3
项目级别
省(略)
研究期限
未(略) (略)
受资助金额
1(略).(略)元(略)
学科
资(略)—(略)
学科代码
未(略)
基金类别
应(略)/(略)助
行(略)计(略) (略)云(略) (略)缘(略)取(略);(略)学(略) (略)动(略)运(略)器(略)流(略)功(略) (略)i(略)E(略) (略)p(略)n(略)E(略);(略)n(略)l(略)o(略)S(略)i(略);(略)g(略)l(略) (略)e(略)S(略)i(略);(略)e(略)e(略)i(略)S(略)i(略)P(略)f(略) (略)E(略)o(略)a(略)a(略) (略)f(略) (略)l(略) (略)c(略)n(略)F(略)
参与者
赖(略)
参与机构
未(略)
项目标书摘要:随着(略)oT)应用的迅速发(略)面临着高延迟、低频(略)型通讯的严峻挑战。(略)展之网路通讯技术正(略)网路边缘计算之趋势(略)种行动边缘计算技术(略)提高频谱效率之B5(略)构,并支援大规模机(略)规划之B5G行动边(略)Central C(略)Edge Clou(略)yer,实作B5G(略)MEC Fog G(略)单一装置及区域的运(略)处理能力,缓和传统(略)所遭遇之资源不足及(略)问题。Edge C(略) Layer之ME(略)ay运用Centr(略)yer之巨量资料深(略)学习模型进行深度学(略)特徵资讯形成大数据(略)两部分,作为独立训(略)Central C(略)巨量资料深度学习平(略)ud Access(略)Fog Gatew(略)习模型可自动调整类(略)及偏移量,使得计算(略)小值。MEC Fo(略)测试资料集作为输入(略)路节点权重值与偏移(略)测流量特徵,藉此验(略)。此外,本计划藉由(略)算智慧学习网路环境(略)Gateway协同(略)分流机制,以无线网(略)与流量卸载功能模组(略) MEC Fog (略)习特性,即时分析M(略)way执行效能以配(略)EC Fog Ga(略)略达到最佳运作效益(略)迟率及减少整体服务(略)G行动边缘雾运算网(略)
Applicati(略): With th(略)elopment (略)nternet a(略) of Thing(略)cations,t(略)nal centr(略)d computi(略)e severe (略)f high la(略)pectral e(略)nd non-ad(略)ine type (略)on.The em(略)ork commu(略)chnology (略) from the(略)d cloud c(略) the netw(略)mputing f(略)ng these (略)The proje(略)plores th(略)rent oper(略)e Computi(略)ologies.W(略)l of redu(略) and spec(略)ency and (略) large-sc(略) type com(略) architec(略)project d(略) B5G Mobi(略) Network (略)e.The pro(略)obile Edg(略)rk archit(略)udes the (略)ud Layer (略)e Cloud A(略).The B5G (略)eway is i(略)to enhanc(略)tational (略)of a sing(略)nd area,s(略)urces and(略)ssing cap(略)o ease th(略)he tradit(略) computin(略)cture suf(略)roblem of(略)nt resour(略) Fog Gate(略)e machine(略)odel esta(略)the deep (略)atform to(略)aracteris(略)tion from(略)types of (略)n the tra(略)ernet div(略)wo parts (略)ent train(略)t and tes(略)The deep (略)atform in(略)g dataset(略)e Edge Cl(略)Layer MEC(略)y.The lea(略) automati(略)ts weight(略)t between(略)work node(略)og Gatewa(略) the test(略) input to(略)t the pre(略)el by cal(略)ights and(略)om the tr(略)l network(略)lly,this (略)sents the(略)ive Manag(略)ce(CMS)ba(略)Fog Gatew(略)izing B5G(略)e Fog Com(略)lligence (略)twork env(略)e propose(略)ervice Tr(略)sion mech(略)contains (略)s Network(略)dule(RNIS(略)c Offload(略)odule(TOF(略)og Gatewa(略)the analy(略)ution per(略)stantaneo(略)figure th(略)ding traf(略)on strate(略)the Deep (略)ature.Eve(略) MEC Fog (略)ieves the(略)erational(略)nd reduce(略)ll networ(略)nd the to(略)loss rate(略),this pro(略)rovide th(略)uality of(略)bile Netw(略)g Computi(略)ent.
项目受资助省
台(略)
- (略)