【摘要】：該文設計了一種面向NDN中名字查找的哈希布魯姆過濾器(HBF)。HBF由位于片內存儲器中的g個計數(shù)器布魯姆過濾器(CBF)、g個計數(shù)器和位于片外存儲器中的g個哈希表組成,每個哈希表與1個CBF和1個計數(shù)器關聯(lián)。為了避免因部分CBF存入名字過多而導致HBF的高誤判率,HBF通過二次哈希選擇算法將NDN路由器中FIB/CS/PIT表項完整信息均勻分散保存于g個CBF和g個哈希表中,同時也利于數(shù)據(jù)包轉發(fā)的并行處理。理論分析和實驗結果表明在名字查找過程中,HBF利用片內存儲器中CBF的定位與過濾作用,大幅度減少片外存儲器的訪問開銷,提高數(shù)據(jù)包轉發(fā)速率,有效避免泛洪攻擊。
[Abstract]:In this paper, a hash Bloom filter (HBF). HBF) for name lookup in NDN is designed, which consists of g counter in on-chip memory, (CBF) filter g counter and g hash table in off-chip memory. Each hash table is associated with 1 CBF and 1 counter. In order to avoid the high error rate of HBF caused by the excessive number of HBF stored in part of the CBF, the complete information of FIB/CS/PIT table in NDN router can be evenly distributed in g CBF and g hash table by using the quadratic hash selection algorithm. At the same time, it also facilitates the parallel processing of packet forwarding. The theoretical analysis and experimental results show that in the process of name searching, the location and filtering function of CBF in on-chip memory can be used to greatly reduce the access cost of out-of-chip memory, improve the packet forwarding rate, and effectively avoid flooding attacks.
【作者單位】： 湖南大學信息科學與工程學院;
【基金】：國家973項目(2012CB315805) 國家自然科學基金(61173167,61472130)
【分類號】：TP393.0

【相似文獻】

相關期刊論文 前1條

1 張震;汪斌強;陳庶樵;郭通;;幾何布魯姆過濾器的設計與分析[J];電子學報;2012年09期

相關碩士學位論文 前1條

1 張生華;一種面向DPI的內存高效的布魯姆過濾器研究[D];湖南大學;2009年

，

本文編號：2198317