http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
NDN에서의 이름 검색을 위한 레벨 우선순위 트라이를 사용한 이중 블룸필터
변하영(Ha-Young Byun),임혜숙(Hyesook Lim) 대한전자공학회 2016 대한전자공학회 학술대회 Vol.2016 No.11
Network traffic has rapidly been increasing in recent years because of massive growth in mobile applications and video streaming services. Named data networking(NDN) technology was proposed as one of the future Internet architectures to handle effectively the ever-increasing network traffic. In order to realize NDN, a high-speed lookup architecture for a forwarding information base (FIB) is essential. This paper proposes a level-priority trie (LPT) and a 2-stage Bloom filter architecture implementing the LPT. The proposed 2-stage Bloom filter architecture is implemented with internal memories only and completes the name lookup without accessing an external memory. Hence, the proposed structure achieves high-speed name lookup performance for the FIB.
변하영(Ha-Young Byun),이정원(Jung-Won Lee),임혜숙(Hyesook Lim) 대한전자공학회 2016 대한전자공학회 학술대회 Vol.2016 No.6
Counting Bloom filters (CBFs) have been popularly used in many applications for the membership queries of dynamic sets, since CBFs can provide delete operations. A CBF uses an array of c-bit counters. The c should be chosen large enough to avoid overflows, and 4 bits are known to suffice for most applications. However, the CBF composed of 4-bit counters wastes memory spaces by allocating 4 bits for every counter, even though around the half of the counter values are zeros in an optimally programmed Bloom filter. In this paper, we propose a simple alternative of a CBF named ternary Bloom filter (TBF). When the TBF consumes the same amount of memory space as a CBF, it is shown that the TBF provides a better false positive rate than the CBF.
이청령(Qingling Li),변하영(Ha-Young Byun),임혜숙(Hyesook Lim) 대한전자공학회 2017 대한전자공학회 학술대회 Vol.2017 No.6
Internet Protocol (IP) address lookup is one of the most challenging tasks for packet forwarding. We implemented a high speed and flexible architecture for IP address lookup with a FPGA. Our IP address lookup architecture is based on a vectored Bloom filter, which returns the output port of the longest matching prefix. The Bloom filter is a space-efficient data structure that can be stored in an on-chip memory. Our implementation shows the average speeds for 1k and 5k are 42 and 68 cycles, respectively.