iEMSs 2008:International Congress on Environmental Modelling and Software Integrating Sciences and Information Technology for Environmental Assessment and Decision Making 4th Biennial Meeting of iEMSs, http://www.iemss.org/iemss2008/index.php?n=Main.Proceedings M. S` anchez-Marr` e, J. B´ ejar, J. Comas, A. Rizzoli and G. Guariso (Eds.) International Environmental Modelling and Software Society (iEMSs), 2008 A taxonomy of robustness in sensor service networks Kym S. Watson and Thomas Usl ¨ ander a a Fraunhofer Institute IITB, Fraunhoferstr. 1, 76131 Karlsruhe, Germany (kym.watson| thomas.uslaender@iitb.fraunhofer.de) Abstract: We consider which aspects of a sensor service network are important in achieving robustness, i.e. the ability to adapt or react to changes and external influences. The sensor service network is viewed as a collection of functional domains ranging from the Sensor Domain up to the Application and User Domains. The characteristics of robustness relate not only to resilience to communication problems, but also to dynamic sensors and services and to the overall system architectural design. Keywords: Robustness, sensor, service network 1 INTRODUCTION The word robust stems from the Latin robustus meaning oaken (like oak wood) or hardy and strong. A widespread oak tree is the Quercus robur, commonly called English oak. Robustness implies resilience and the ability to adapt to changes and external influences. The degree of robustness is determined by the types of changes and influences which can be coped with. The attribute robust is broadly applicable to practically all entities, including living beings, natural and man-made objects, processes and technical systems. Ideally, a robust object or system has its own inherent ability to react and / or adapt to changes and external influences, even to those for which no explicit reaction has been foreseen. The workshop contribution will propose a classification of factors characterizing robustness in a sensor service network. The essential issues of what should be made robust to which changes and influences are summarized in the sections below. 2 FUNCTIONAL SCOPE OR WHAT MUST BE ROBUST The complete functionality of a sensor service network can be structured into domains as shown in Figure 1. Layer-like interfaces are not implied between the domains, instead the domains should be considered as a loose functional structure. The Sensor Domain covers the actual sensor or transducer technology, and the communication within the sensor network (wired or wireless such as ZigBee) and to gateway nodes with access to higher level networks, which are typically IP based. The sensor nodes may be mobile and belong to varying sub-networks. Information sources such as a database of archived data or a model- based algorithm can both be considered as sensors as defined in the Observation & Measurement Model of OGC, cf. Cox [2007]. The overlying functional domains serve to find, integrate, generate and present information orig- inating in the Sensor Domain. The overall goal is of course to achieve robustness in the User