Overview

The project will evaluate the applicability of compositional calculi for estimating quality of service (QoS) properties arising in the development of an e-Science infrastructure such as the Grid. The properties we are investigating include provenance, accuracy, reliability and run time. The goals of the project are to define and represent QoS properties, to formulate QoS calculi, to embed the calculi in some framework to form an application system, and to do evaluation on the application system.

Grant Information

EPSRC grant GR/S62949 (R37343) (funded by CS for e-Science Programme). Duration: 1 Oct 2003 - 30 Sept 2006. PhD Project Student: Lin Yang. Grant Holders: Alan Bundy and Stuart Anderson. Co-supervisors: Dave Berry, Sophie Huczynska and Conrad Hughes.

Current Directions

We are developing a Bell-Curve Calculus to describe some quantifiable QoS properties, including run time, accuracy and reliability (mean time to fail). After completing the calculus, we will apply it to some real grid environment and do evaluation and adjustment on it.

Publications

• Towards a Bell-Curve Calculus and its Application to e-Science (PPT) by Lin Yang, in the CIAO (CLAM/INKA/OMRS) 2005 Workshop, 4th-6th April 2005, Nottingham, UK. (CIAO-05.ppt)
• Towards a Bell-Curve Calculus and its Application to e-Science (PDF) by Lin Yang, CS poster, presented in EPSRC e-Science Meeting 2005, 21st April 2005, NeSC, Edinburgh, UK. (EPSRC2005.pdf)
• Towards a Bell-Curve Calculus and its Application to e-Science (PDF) by Lin Yang, poster + poster talk, presented in the ARW (Automated Reasoning Workshop), 29th July 2005, Edinburgh, UK. (ARW2005.pdf)
• Towards a Bell-Curve Calculus and its Application to e-Science (PDF) by Lin Yang, Alan Bundy, Dave Berry, Conrad Hughes, in the Workshop on Grid Performability Modelling and Measurement, March 2006, NeSC, Edinburgh, UK. (NeSC495.pdf)
• Towards a Bell-Curve Calculus for e-Science (PDF) by Lin Yang, Alan Bundy, Dave Berry, Conrad Hughes, in UK e-Science All Hands Meeting 2006, 18th-21st September 2006, Nottingham, Edinburgh, UK. (AHM2006.pdf)

Related Projects

• The Deductive Synthesis of Grid Workflows for e-Science aims to evaluate the applicability of deductive synthesis to for automated rapid and customised development of Grid applications.
• PASOA (Provenance Aware Service Oriented Architecture) aims to investigate the concept of provenance and its use for reasoning about the quality and accuracy of data and services in the context of eScience. They have established a provenance system, which we would include our Bell-Curve Calculus in.
• DIGS (Dependability Infrastructure for Grid Services) is to investigate in fault-tolerance system and other quality of service issues in service-oriented architectures. They developed a system called Agrajag to implement some operations and measurements on some basic models of stochastic distributions. It is a basic tool we use to validate our Bell-Curve Calculus in an experimental way.
• myGrid developed a middleware for in silico experiments in biology, in which the concepts and structures concerned with provenance provided some inspiration.