ArticleDirectory
- [Thesis 1] A Probabilistic Approach to modeling and estimating the QoS of web-services-based workflows (is, 2007)
- [Thesis 2] QoS prediction for composite Web Services with transactions (icsoc workshops, 2006)
[Thesis 1] A Probabilistic Approach to modeling and estimating the QoS of web-services-based workflows (is, 2007)
San-Yih Hwang, haojun Wang, Jian Tang, jaideep Srivastava:A Probabilistic Approach to modeling and estimating the QoS of web-services-based workflows. INF. Sci. 177 (23): 5484-5503 (2007)
Based on the previous article. Propose a probabilistic framework for QoS computation.
A probability-Based QoS model for atomic and composite WSS that views a QoS measure as a discrete random variable with a probability mass function (PMF ),
Three QoS types are discussed: cost, response time, and reliability.
(S2.3) five basic ureured constructs
(1) sequence
(2) Parallel split/join (and Split/and join)
(3) Exclusive Choice (exclusive split/exclusive join)
(4) discriminator (and Split/or join)
When multiple services are executed synchronously, the first returned result is used and then ignored.
(5) loop
S3.2
Calculate the cost/Response Time/reliability of the five basic structures above.
[Thesis 2] QoS prediction for composite Web Services with transactions (icsoc workshops, 2006)
Jiangxia Wu, fangchun Yang:QoS prediction for composite Web Services with transactions. Icsoc workshops 2006: 86-94
This article discusses the calculation of QoS below CWS with the transaction condition in CWS:
Response mence (Response Time), cost, reputation
Assumption:
(1) failure will not happen to two services simultaneously
(2) exception handling operations will not fail
The method is relatively earthy, that is, adding the path with transaction to the flowchart, which is equivalent to a new CWS, as long as the QoS of the CWS is calculated.
Note: The reliability is not estimated in this article. The error rate ETA refers to the deviation between the estimated QoS and the actual QoS.