Use this URL to cite or link to this record in EThOS:
Title: Adaptation for mobile systems
Author: Edmonds, T. M.
Awarding Body: University of Cambridge
Current Institution: University of Cambridge
Date of Award: 2001
Availability of Full Text:
Full text unavailable from EThOS.
Please contact the current institution’s library for further details.
Mobile devices operate in a environment with limited and volatile resource availability, suffering from poor CPUs, memory and battery limitations and low rate, unreliable communications when compared with their fixed counterparts. Under such conditions, applications for mobile computing require adaptation to make best use of the available resources without assuming the minimum set. This dissertation argues that such adaptation should be integrated, with centralised resource management for efficiency and fairness, but with application specific reconfiguration to ensure that adaptation remains meaningful. Current integrated adaptation systems provide resource change events to applications but fail to say exactly how to react to them. This dissertation presents Pervasive Adaptation as a new approach to integrated adaptation. This approach separates the concerns of adaptation from application functionality by providing automatic dynamic adaptation as a service external to the application. This service instruments and models an application's operation and performs reconfiguration of the application, achieving adaptation through manipulation of the locality and fidelity of application processing and data. Separating the adaptation service provides two key benefits: firstly, it reduces the work required of an application developer and secondly, it allows the same adaptation service to adapt all of the applications running in the system, ensuring cooperation and fairness. The culmination of this work is the design and implementation of the DPROJ framework which provides a complete infrastructure for the construction, deployment and dynamic adaptation of mobile applications. It not only provides the mechanisms for adaptation but also the "know-how". The adaptation service reconfigures the structure of the application based on cost-benefit analysis. Costs are estimated through a resource unification scheme under which many different resources such as CPU, communications and battery can be combined into a common metric. Performance is maintained as a dimensionless, application specific notion. Running in a simulated network environment, DPROJ is demonstrated to successfully manage the adaptation of several applications including media streaming and vision processing.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available