2nd International Workshop on Engineering Emergence in Decentralised Autonomic Systems
to be held at
The 4th IEEE International Conference on Autonomic Computing (ICAC 2007)
Home Call for Papers Important Dates Author Guidelines Publication Channels Committees Program Papers Contact Previous Events
ACM   IEEE and IEEE Computer Society  

Valid HTML 4.01!Valid CSS!

Please upgrade to the latest browser version if the site doesn't show correctly.

Latest News

  • 07-06-2007 A schedule or program is online!
  • 17-04-2007 The list of accepted papers is put online
  • 16-04-2007 Notifications are sent to authors.
    If you are an author and have not received it, let us know.
  • 23-02-2007 The deadline was extended to March 18, 2007.
  • 22-01-2007 The program committee was finalized, see committees.
  • 21-12-2006 Workshop proposal was accepted by ICAC.

Introduction

Autonomic Computing addresses the need for systems to cope with and manage complex, highly dynamic, and changing operating conditions autonomously. Furthermore, many modern computing systems effectively operate in a decentralised fashion, adding an extra layer of complexity to the management (or even the prediction) of global system behaviour. Autonomic Computing arguably addresses one aspect of the problem (autonomous operation of individual units) but doesn’t inherently guarantee the emergence of the desired collective behaviour in systems where central or global control is impossible for one or several of the following reasons:

  • The information needed to make decisions is inherently localised and cannot be gathered centrally (e.g. ad-hoc networks).
  • A highly dynamic context implies that, even when information is gathered centrally, it is obsolete when it reaches the manager.
  • The amount of processing required to orchestrate the system operation is beyond the capability of any participating unit.
  • The centrally computed global solution cannot be practically implemented.
  • Short response time is more critical than efficiency (quick, local reaction to events is better than a globally optimised but delayed solution).

Self-organising emergence is an approach to engineering decentralised systems so as to make them capable of dynamically adapting to changes without external intervention. The global functionality dynamically arises from the local autonomous decisions and interactions between individual entities based on local information usually communicated through local channels (long-range diffusion being typically achieved via “gossiping”). These entities are not explicitly aware of the resulting global behaviour. The lack of global or central control implies the need to design decentralised coordination mechanisms and the local rules governing individual behaviour with the explicit objective to take full advantage of emergent properties.

The ultimate challenge in engineering fully decentralised autonomous systems is to find a disciplined approach to foster globally coherent and desired system behaviour. Many research issues are still open, mainly because of the lack of a clear step-plan to map the desired global behaviour onto a specific set of local decision and interaction rules. As a result, exploitation of emergent behaviour in system design is still in its infancy. This workshop aims at gathering research results that could contribute to addressing this challenge.

Webmaster: Tom De Wolf