科研信息化技术与应用
科研信息化技術與應用
과연신식화기술여응용
E-science Technology & Application
2015年
2期
28-36
,共9页
冯宗明%李俊%吴海博%智江
馮宗明%李俊%吳海博%智江
풍종명%리준%오해박%지강
内容中心网络%缓存发现%控制平面%选择存储
內容中心網絡%緩存髮現%控製平麵%選擇存儲
내용중심망락%완존발현%공제평면%선택존저
Content-Centric Network (CCN)%cache discovery%control plane%selective storage
内容中心网络(Content-Centric Network, CCN)是未来互联网的重要发展方向.网内缓存是 CCN网络的重要特征,对 CCN内容传输性能具有重要影响.网内缓存的内容发现效率与网内缓存性能密切相关.传统 CCN网络缓存发现方法是通过请求包在数据平面转发,沿途机会性地命中缓存实现的,具有一定的随机性、盲目性,可能导致缓存内容无法被高效利用.本文提出一种在控制平面解决缓存可用性的方法,结合拓扑、缓存容量以及用户请求分布计算出"值得"缓存的内容进行存储,同时将其向外通告,使其参与路由计算,以便后续请求快速准确地发现并利用缓存内容.实验结果表明,本文方法可使缓存命中率提高 20%左右,服务器负载降低 15%左右.
內容中心網絡(Content-Centric Network, CCN)是未來互聯網的重要髮展方嚮.網內緩存是 CCN網絡的重要特徵,對 CCN內容傳輸性能具有重要影響.網內緩存的內容髮現效率與網內緩存性能密切相關.傳統 CCN網絡緩存髮現方法是通過請求包在數據平麵轉髮,沿途機會性地命中緩存實現的,具有一定的隨機性、盲目性,可能導緻緩存內容無法被高效利用.本文提齣一種在控製平麵解決緩存可用性的方法,結閤拓撲、緩存容量以及用戶請求分佈計算齣"值得"緩存的內容進行存儲,同時將其嚮外通告,使其參與路由計算,以便後續請求快速準確地髮現併利用緩存內容.實驗結果錶明,本文方法可使緩存命中率提高 20%左右,服務器負載降低 15%左右.
내용중심망락(Content-Centric Network, CCN)시미래호련망적중요발전방향.망내완존시 CCN망락적중요특정,대 CCN내용전수성능구유중요영향.망내완존적내용발현효솔여망내완존성능밀절상관.전통 CCN망락완존발현방법시통과청구포재수거평면전발,연도궤회성지명중완존실현적,구유일정적수궤성、맹목성,가능도치완존내용무법피고효이용.본문제출일충재공제평면해결완존가용성적방법,결합탁복、완존용량이급용호청구분포계산출"치득"완존적내용진행존저,동시장기향외통고,사기삼여로유계산,이편후속청구쾌속준학지발현병이용완존내용.실험결과표명,본문방법가사완존명중솔제고 20%좌우,복무기부재강저 15%좌우.
As a promising direction of future Internet, Content-Centric Network (CCN) has attracted world-wide attention. Caching is an important feature of CCN network, which has significant impacts on the performance of content transmission. The efficiency of cached content discovery has a close relationship with the caching performance. In traditional methods of cache discovery in CCN, requests are forwarded in the data plane and caches are hit along the path in an opportunistic manner. This approach has the characteristic of randomness and blindness, which may make the cache content not be used efficiently. This paper presents a new method to achieve the availability of caching in the control plane, which utilizes topology, cache capacity and user requests distribution to calculate the worthy cache contents and store them. Meanwhile, the corresponding cache contents are announced and participate in the routing calculation, so that the subsequent requests can quickly and accurately find and exploit these cache contents. The experimental results show that this method can improve the rate of cache hit by about 20% and reduce the server load by about 15%.