Autonomic Computing: Principles, Design and Implementation (Undergraduate Topics in Computer Science)

Autonomic Computing: Principles, Design and Implementation (Undergraduate Topics in Computer Science)

Philippe Lalanda, Ada Diaconescu

Language: English

Pages: 288

ISBN: 1447150066

Format: PDF / Kindle (mobi) / ePub

This textbook provides a practical perspective on autonomic computing. Through the combined use of examples and hands-on projects, the book enables the reader to rapidly gain an understanding of the theories, models, design principles and challenges of th

















unpredictability is relative here; even the most autonomous system requires some programming of what to do when failure is detected. Let us illustrate this by returning to our audio-server example. Recall that the purpose of this system was to adapt its audio encoding code depending on how it perceived the link between the audio server and the user at a moment in time. The overall goal is that there should never be a moment of audio silence during playback. To test how well this system is able

alleviating these challenges. Notably such approaches can draw inspiration from fields such as multi-­agent systems, proposing various techniques for agent collaboration or competition, and self-organising systems, studying the opportunistic self-assembly of systems from simpler elements. Finally, multi-criteria optimisation techniques can be adopted to manage multiple and possibly conflicting goals. 10.3.3 Who Guards the Guards? Trust and Assurances in Autonomic Computing When computing moved

destination. AutoNav has been successfully tested for guiding the Deep Space 1 spacecraft (1999) and Deep Impact (2005). 24NASA’s Mars Science Laboratory (MSL) mission: 25ESA: European Space Agency— 26Mars Sample Return mission: 27NASA’s Autonomous Nano Technology Swarm (ANTS) program— 28NASA’s ANTS Prospecting Asteroid Mission (PAM), expected timeframe:

problems. Autonomic computing can be included in this category if one considers that the particular problems to solve, as specified by human administrators, centre around the administration of software systems. Within this context, autonomic computing aspires to a considerably more modest objective than AI. The minimum level of intelligence required will depend on the complexity of the targeted computing system to administer and of its execution environment. This implies that rather than creating

field. 4.4.2 Hierarchical Versus Decentralised Organisation An approach to multi-agent cooperation is a hierarchical structuring of agents [25]. This organisation can be advantageously applied to the design of autonomic systems, as illustrated by Fig. 4.11. Autonomic elements are organised into a hierarchy where elements can set goals to the elements of lesser level, which, in turn, provide feedback about their behaviour. Fig. 4.11Hierarchical organisation Hierarchical autonomic management

Download sample