Demonstration venue · read-only. Every page can be browsed; the buttons that would change it are switched off. Create an account to run TaxoReview on your own data.

Victor T.-S. Shi

dblp:16/4156 · DBLP profile ↗
← Back
2ranked-venue papers
1as first author
0since 2021 · last 1999
—ORCID · none

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

Computer networks · 1 · 1 first-authorDatabases, data management, data science and information retrieval · 1

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
2 papers
Network performance modeling · 75% Internet architecture and protocols · 9% Software-defined and programmable networks · 9%
Computer architecture, parallel and distributed computing, and storage systems
1 paper
Distributed systems · 100%

Topics — the 8 heaviest of 8, each with the papers that count most for it

TopicWeightPapersLastEvidence papers
Distributed systems › distributed database
commit protocol
0.011999
Atomic Commitment in Database Systems over Active Networks · ICDE 1999
Distributed systems › distributed database
distributed transactions
0.011999
Atomic Commitment in Database Systems over Active Networks · ICDE 1999
Network performance modeling › loss systems
blocking probability
0.011998
The performance of a link with multipriority traffic · IEEE Trans. Commun. 1998
Network performance modeling › queueing analysis
priority queueing
0.011998
The performance of a link with multipriority traffic · IEEE Trans. Commun. 1998
Network performance modeling
queueing analysis
0.011998
The performance of a link with multipriority traffic · IEEE Trans. Commun. 1998
Internet architecture and protocols › future internet architecture
active networks
0.011999
Atomic Commitment in Database Systems over Active Networks · ICDE 1999
Software-defined and programmable networks › programmable data plane
in-network computation
0.011999
Atomic Commitment in Database Systems over Active Networks · ICDE 1999
Wireless networking › wireless link
wireless link performance
0.011998
The performance of a link with multipriority traffic · IEEE Trans. Commun. 1998

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

blackboard-based voting · 0.0active node coordination · 0.0steady-state balance equations · 0.0markov analysis · 0.0
YearPublicationVenuePosition
1999 Atomic Commitment in Database Systems over Active Networks
abstract
Traditional communication networks have functioned only to deliver information packets from one end system to another. As the cost of computing power continues to decrease, it is worthwhile to consider the benefits of adding more of that computing power to the network itself, either to enhance services or to trade off against other costs such as time, bandwidth and storage. This new area of research is called Active Networking. Active networking represents a new approach in which network switches or nodes perform customized computing as the messages flowing through them. This paper proposes a novel approach to atomic commitment of transactions in distributed database systems. In our approach, active nodes participate in the process of atomic commitment of distributed transactions by maintaining passive "blackboards" which record voting status of participant subtransactions. One active node is delegated the coordinator function so that when rendezvous is there, commit decisions can be made. Since the decision is made from an internal network node (selected to be close to the participant nodes), the uncertainty time delay can be significantly reduced. This is especially critical in wide area networks. Other active nodes can be used to suppress duplicated votes sent in the upstream direction. Additionally, once a negative vote is received at any active node, children participants can be notified of the abort decision immediately and therefore much earlier than would otherwise be possible. This approach is shown to yield performance improvements over traditional methods, in terms of delay and network message traffic.
William Perrizo, Victor T.-S. Shi
ICDE3
1998 The performance of a link with multipriority traffic
abstract
This letter discusses the performance of a link carrying calls of preemptive priorities. General /spl kappa/-dimensional steady-state balance equations are deduced and the computational complexity of two solution methods are compared through an example of three priorities. The performance of the link is studied through two conventional metrics for blocking probabilities and preempting probabilities. The "peaking" phenomenon of the preempting line discussed can be used in the design of switches and networks. The coincidence of the preempting probability line with the blocking probability line, displayed in the discussion of numerical results, suggests that better metrics for preemptive performance must be found.
Victor T.-S. Shi, Wang Chu, William Perrizo
IEEE Trans. Commun.1