Keqin Zhu

dblp:85/5334 · DBLP profile ↗
← Back
2ranked-venue papers
0as first author
0since 2021 · last 2006
—ORCID · none

Domains — the database's venue-derived domains; a paper can count in several

Computer networks · 2

Expertise — from the expertise taxonomy: the topics of the expert's papers under the CCF categories. A weight counts papers with recency: 1 for a paper about the topic, 0.3 when the topic is its context, halved every five years.

Computer networks
1 paper
Network management and operations · 67% Internet architecture and protocols · 33%
Theoretical computer science
1 paper
Automata and formal languages · 100%

Topics — the 5 heaviest of 5, each with the papers that count most for it

TopicWeightPapersLastEvidence papers
Network management and operations › network testing
protocol conformance testing
0.011993
Optimal length test sequence generation using distinguishing sequences · IEEE/ACM Trans. Netw. 1993
Internet architecture and protocols › protocol engineering
protocol testing
0.011993
Optimal length test sequence generation using distinguishing sequences · IEEE/ACM Trans. Netw. 1993
Network management and operations › network testing › protocol conformance testing
test sequence generation
0.011993
Optimal length test sequence generation using distinguishing sequences · IEEE/ACM Trans. Netw. 1993
Automata and formal languages › finite automata
distinguishing sequences
0.011993
Optimal length test sequence generation using distinguishing sequences · IEEE/ACM Trans. Netw. 1993
Automata and formal languages
finite automata
0.011993
Optimal length test sequence generation using distinguishing sequences · IEEE/ACM Trans. Netw. 1993

Methods — techniques the papers use, named apart from their topics

overlapping test segments · 0.0d-method · 0.0
YearPublicationVenuePosition
2006 Distributed delay constrained multicast routing algorithm with efficient fault recovery
abstract
Abstract Existing distributed delay constrained multicast routing algorithms construct a multicast tree in a sequential fashion and need to be restarted when failures occur during the multicast tree construction phase or during an on‐going multicast session. This article proposes an efficient distributed delay constrained multicast routing algorithm that constructs a multicast tree in a concurrent fashion by taking advantage of the concurrency in the underlying distributed computation. The proposed algorithm has a message complexity of O(mn) and time complexity of O(n) in the worst case, wheremis the number of destinations andnis the number of nodes in the network. It constructs multicast trees with the same tree costs as the ones constructed by well‐known algorithms such as DKPP and DSHP while utilizing 409 to 1734 times fewer messages and 56 to 364 times less time than these algorithms under comparable success rate ratios. The proposed algorithm has been augmented with a fault recovery mechanism that efficiently constructs a multicast tree when failures occur during the tree construction phase and recovers from any failure in the multicast tree during an on‐going multicast session without interrupting the running traffic on the unaffected portion of the tree. © 2005 Wiley Periodicals, Inc. NETWORKS, Vol. 47(1), 37–51 2006
Hasan Ural, Keqin Zhu
Networks2
1993 Optimal length test sequence generation using distinguishing sequences
abstract
The optimization of the length of test sequences for finite state machine based protocol conformance testing is studied. The study focuses on test generation methods, called D-methods, that utilize distinguishing sequences in the construction of test segments. The extent of the optimization of the length of a test sequence is investigated with respect to two cases. The first case establishes the lower bound for the length of test sequences generated by any D-method that overlaps test segments. The second case establishes the lower bound for the length of test sequences generated by any D-method that does not overlap test segments. It is observed that the reduction in the length of test sequences due to overlapping is significant. An efficient algorithm for the generation of test sequences is proposed. This algorithm utilizes a distinguishing sequence and overlaps test segments. Sufficiency conditions are given both for finding a minimum- length test sequence in polynomial time and for constructing the optimal length test sequences by this algorithm.>
Hasan Ural, Keqin Zhu
IEEE/ACM Trans. Netw.2