Please use this identifier to cite or link to this item:
Type: Thesis
Title: Intelligent discovery, configuration and composition of devices in a distributed system.
Author: Kaiyan, Nicole
Issue Date: 2014
School/Discipline: School of Computer Science
Abstract: Establishing access to the functionality of input/output devices across a distributed system presents significant challenges worth investigating. To accomplish this requires locating devices and understanding their identity so that requests for them can be satisfied. Current systems require the domain of requestable devices be resolved beforehand and that they be identified as discrete items. Additionally, configuring devices only happens if an operating system has access to suitable drivers and composition is conducted by middleware applications, without any system service awareness. These approaches restrict device use in a distributed context and fail to provide a satisfactory solution. This research has the goal of accomplishing access to devices in a distributed system without such constraints. We present an approach where devices are described by a language based upon a rich taxonomy of form and function. Requests for devices are formulated using the same language and a matching process is employed to satisfy requests. The devices capable of being matched may be rich in functionality, complex, consist of sub-units, and include those yet to be developed. A taxonomy capturing the scope of form and function populates the description space with terms relevant to devices. The description space is structured hierarchically to manage complexity. A contribution is also made to improving the design and integration of operating systems components, in particular, those services responsible for managing devices, from configuration through to composition, and accomplishing such across a distributed system. In this context, configuration serves as a local system response to device connections, ensuring they become operational then advertising their availability remotely. Distributed composition services receive these notifications, adding device descriptions to a database for use when matching requests. We have adopted the language Prolog to describe devices and for implementing a distributed system. It supports a database of device descriptions in the form of assertions and provides powerful support for matching via an inference engine. The inference engine systematically examines a potentially large and complex search space for an acceptable extent of correspondence. Requests can be expressed minimally as those elements of a device relevant to matching. A consequence of this style of matching is the ability to allow requesters partial access to device functionality as a result of incomplete satisfaction. Through awareness of the device domain, the composition process handles allocation of control and access arbitration. A demonstration of structural matching is provided through a fully worked example. We set out to build a distributed system with sufficient capability to investigate structural matching between requests for functionality and device identities. Furthermore, to accomplish this dynamically when devices connect and permit access to their functionality as the outcome of matching requests. The resultant schema presents a comprehensive solution by combining a structured language, expressive enough to describe current devices and future possibilities, with a tailored inference engine, designed to compose an entire distributed system.
Advisor: Vaughan, Francis Alexander
Barter, Christopher
Dissertation Note: Thesis (Ph.D.) -- University of Adelaide, School of Computer Science, 2014
Keywords: operating systems; devices; distributed systems; discovery; configuration; composition
Provenance: This electronic version is made publicly available by the University of Adelaide in accordance with its open access policy for student theses. Copyright in this thesis remains with the author. This thesis may incorporate third party material which has been used by the author pursuant to Fair Dealing exceptions. If you are the owner of any included third party copyright material you wish to be removed from this electronic version, please complete the take down form located at:
Appears in Collections:Research Theses

Files in This Item:
File Description SizeFormat 
01front.pdf107.86 kBAdobe PDFView/Open
02whole.pdf6.05 MBAdobe PDFView/Open
  Restricted Access
Library staff access only256.91 kBAdobe PDFView/Open
  Restricted Access
Library staff access only6.24 MBAdobe PDFView/Open

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.