1 |
Formal modelling and analysis of dynamic reconfiguration of dependable systemsBhattacharyya, Anirban January 2013 (has links)
The contribution of this thesis is a novel way of formally modelling and analyzing dynamic process reconfiguration in dependable systems. Modern dependable systems are required to be flexible, reliable, available and highly predictable. One way of achieving flexibility, reliability and availability is through dynamic reconfiguration. That is, by changing at runtime the structure of a system – consisting of its components and their communication links – or the hardware location of its software components. However, predicting the system’s behaviour during its dynamic reconfiguration is a challenge, and this motivates our research. Formal methods can determine whether or not a system’s design is correct, and design correctness is a key factor in ensuring the system will behave predictably and reliably at runtime. Therefore, our approach is formal. Existing research on software reconfiguration has focused on planned reconfiguration and link mobility. The focus of this thesis is on unplanned process reconfiguration. That is, the creation, deletion and replacement of processes that is not designed into a system when it is manufactured. We describe a process algebra (CCS<sup>dp</sup>) which is CCS extended with a new type of process (termed a fraction process) in order to model process reconfiguration. We have deliberately not introduced a new operator in <sup>dp</sup> in order to model unplanned reconfiguration. Instead, we define a bisimulation (~<sub>of</sub>) that is used to identify a process for reconfiguration by behavioural matching. The use of behavioural matching based on ~<sub>of</sub> (rather than syntactic or structural congruence-based matching) helps to make models simple and terse. However, ~<sub>of</sub> is too weak to be a congruence. Therefore, we strengthen the conditions defining ~<sub>of</sub> to obtain another bisimulation (~<sub>dp</sub>) which is a congruence, and (therefore) can be used for equational reasoning. Our notion of fraction process is recursive to enable fractions to be themselves reconfigured. We bound the depth of recursion of a fraction and its successors in order to ensure that ~<sub>of</sub> and ~<sub>dp</sub> are decidable. Furthermore, we restrict the set of states in a model of a system to be finite, which also supports decidability of the two bisimulations and helps model checking. We evaluate CCS<sup>dp</sup> in two ways. First, with respect to requirements used to evaluate other formalisms. Second, through a simple case study, in which the reconfiguration of an office workflow is modelled using CCS<sup>dp</sup>.
|
2 |
Stability and responsibility : visualising dependencies in large software systemsCain, James Westland January 2002 (has links)
No description available.
|
3 |
Software outsourcing vendors' readiness model (SOVRM)Khan, Siffat Ullah January 2011 (has links)
CONTEXT - Offshore software development outsourcing (OSDO) is a modern business strategy for developing high quality software in low-wage countries at low cost. OSDO is a contract-based relationship between client and vendor organisations in which a c1ient(s) contracts out all or part of its software development activities to a vendor(s), who provides agreed services in return for remuneration. Vendor's readiness plays an important role in the successful outcomes of OSDO projects. OBJECTIVE - The objective of this thesis is to develop a software outsourcing vendors' readiness model (SOVRM) to help vendors for assessing and improving their readiness for OSDO activities. The SOVRM is primarily developed to assist OSDO vendor organisations. However, it is also beneficial to client organisations as client organisations can take benefit from this research by getting to know about the areas in which vendors can be assessed based on their own priorities. Moreover, clients can make a better informcd decision of their choice of OSDO vendors. In developing the SOVRM model, we considered other similar efforts addressing similar areas, specifically CMMI and ISO 9001. SOVRM has its distinctive and unique features that differentiate it from other models. Specifically, SOVRM is a comprehensive model focussing on assisting outsourcing companies in assessing their readiness for software development outsourcing. It is not a software process improvement model or standard such as CMMI and ISO 9001. Vendor companies which are CMMI or ISO 900 I certified may not be ready for software development outsourcing as the characteristics of CMMI and ISO 9001 is to improve the software development capabilities of companies instead of improving their outsourcing readiness. However, we included CMMI or ISO 900 I certification as one of the success factors defined in SOVRM.
|
4 |
Exploiting the architectural characteristics of software components to improve software reuseAlkazemi, Basem Yousef January 2009 (has links)
Software development is a costly process for all but the most trivial systems. One of the commonly known ways of minimizing development costs is to re-use previously built software components. However, a significant problem that source-code re-users encounter is the difficulty of finding components that not only provide the functionality they need but also conform to the architecture of the system they are building. To facilitate finding reusable components there is a need to establish an appropriate mechanism for matching the key architectural characteristics of the available source-code components against the characteristics of the system being built. This research develops a precise characterization of the architectural characteristics of source-code components, and investigates a new way to describe how appropriate components for re-use can be identified and categorized.
|
5 |
Combining control executives and software specificationsTreharne, Helen Eleri January 2000 (has links)
No description available.
|
6 |
Dynamic configuration of distributed multimedia componentsMitchell, Robert Scott January 2000 (has links)
No description available.
|
7 |
A distributed real-time operating system for a multi-processor StrongARM networkLiew, Kenny Eng Wee January 2002 (has links)
No description available.
|
8 |
An approach to systematic software reuse based on plugging components into an architectureMugisa, Ezra Kaahwa January 2000 (has links)
No description available.
|
9 |
A class of software reliability growth models based on exercise frequenciesAndrew, Alan W. January 1990 (has links)
No description available.
|
10 |
Data representations in high-level programming languagesElder, John Watson George January 1975 (has links)
No description available.
|
Page generated in 0.0469 seconds