SIWN Abstracts Index

sai: cosiwn.2007.09.131

Building Knowledge Networks Using Panoramic Images

Stefano Valtolina, Stefano Franzoni and Pietro Mazzoleni

System and Information Sciences Notes, Vol. 2, No. 1, September 2007, pp. 1-6

Abstract: This paper presents a system in which 360 panoramic images are used to disseminate cultural heritage information by accessing open networks of knowledge. The system has been designed following the patterns of Interaction Design. During the development of the system, two new patterns especially useful to the dissemination of cultural heritage content have been recognized. The first, called “Knowledge Network”, offers a solution to the problem of re-contextualizing collections of artifacts according to a given theme the user can choose. The second, called “Virtual Visit”, guides the development of a seamless virtual space composed by a network of panoramic images. The two patterns can be naturally combined: the items organized by Knowledge Network can be displayed to the user in a virtual exhibition arranged by Virtual Visit. In this paper is argued how this approach can facilitate the development of applications easily customizable by the user and characterized by a high level of interactivity.

Keywords: cultural heritage, interaction patters, panoramic images.

sai: cosiwn.2007.09.132

The Role of Reliability, Availability and Serviceability Models in the Design and Evaluation of Self-Healing Systems

Rean Griffith, Ritika Virmani and Gail Kaiser

System and Information Sciences Notes, Vol. 2, No. 1, September 2007, pp. 7-13

Abstract: In an idealized scenario, self-healing systems predict, prevent or diagnose problems and take the appropriate actions to mitigate their impact with minimal human intervention. To determine how close we are to reaching this goal we require analytical techniques and practical approaches that allow us to quantify the effectiveness of a system’s remediation mechanisms. In this paper we apply analytical techniques based on Reliability, Availability and Serviceability (RAS) models to evaluate individual remediation mechanisms of select system components and their combined effects on the system. We demonstrate the applicability of RAS-models to the evaluation of self-healing systems by using them to analyze various styles of remediations (reactive, preventative etc.), quantify the impact of imperfect remediations, identify sub-optimal (less effective) remediations and quantify the combined effects of all the activated remediations on the system as a whole.

Keywords: Kheiron, RAS-models, self-healing evaluation.

sai: cosiwn.2007.09.133

On Expressing and Validating Requirements for the Adaptivity of Self-Organizing Multi-Agent Systems

Jan Sudeikat and Wolfgang Renz

System and Information Sciences Notes, Vol. 2, No. 1, September 2007, pp. 14-19

Abstract: As Multi-Agent Systems (MAS) are composed of organizations of (inter–) acting, autonomous and pro-active agents, they provide means to the construction of self–organizing processes in distributed applications. Engineering approaches to self–organizing dynamics in MAS are supported by simulation–driven development procedures and established decentralized agent coordination mechanisms. In this paper, we examine current approaches to self-organizing MAS development and identify a lack of requirements analysis techniques. As the use of self–organizing mechanisms typically enables continuous MAS reconfiguration, i.e. structural adjustment due to environmental forces, we show how expectations on this MAS adaptivity can be expressed and validated. Corresponding models of intended MAS behaviors consider both the intended system configurations as well as the system dynamics that steer system (re–) configuration. It is argued that self–organizing MAS development typically intends established pattern of dynamics and the specification and validation of these adaptation dynamics is exemplified, transferring software engineering practices to MAS development. (Note 1)

Keywords: adaptivity, decentralized coordination, multi-agent system, requirements engineering, self-organization.

sai: cosiwn.2007.09.134

Towards Policy-Based Self-Configuration of Embedded Systems

Mariusz Pelc and Richard Anthony

System and Information Sciences Notes, Vol. 2, No. 1, September 2007, pp. 20-26

Abstract: This paper describes work towards the deployment of flexible self-management into real-time embedded systems.

A challenging project which focuses specifically on the development of a dynamic, adaptive automotive middleware is described, and the specific self-management requirements of this project are discussed. These requirements have been identified through the refinement of a wide-ranging set of use cases requiring context-sensitive behaviours. A sample of these use-cases is presented to illustrate the extent of the demands for self-management.

The strategy that has been adopted to achieve self-management, based on the use of policies is presented. The embedded and real-time nature of the target system brings the constraints that dynamic adaptation capabilities must not require changes to the run-time code (except during hot update of complete binary modules), adaptation decisions must have low latency, and because the target platforms are resource-constrained the self-management mechanism must have low resource requirements (especially in terms of processing and memory). Policy-based computing is thus an ideal candidate for achieving the self management because the policy itself is loaded at run-time and can be replaced or changed in the future in the same way that a data file is loaded. Policies represent a relatively low complexity and low risk means of achieving self-management, with low run-time costs. Policies can be stored internally in ROM (such as default policies) as well as externally to the system.

The architecture of a designed-for-purpose powerful yet lightweight policy library is described.

A suitable evaluation platform, supporting the whole life-cycle of feasibility analysis, concept evaluation, development, rigorous testing and behavioural validation has been devised and is described.

Keywords: automotive control systems policy-based computing, self-management, middleware.

sai: cosiwn.2007.09.135

Semantic Interoperability in Self-Configuring Service Networks for Context-Driven Decision Making

Alexander Smirnov, Mikhail Pashkin, Nikolay Shilov and Tatiana Levashova

System and Information Sciences Notes, Vol. 2, No. 1, September 2007, pp. 27-32

Abstract: Today the process of complex decision making faces a set of problems during processing of large amount of information from heterogeneous distributed sources. The problems include search of right sources, extraction of their content, integration of the extracted content, and generation of alternative solutions for decision making. The presented approach proposes usage of self-configuring networks of knowledge sources and problem solvers for the above problem solving. The paper is devoted to questions of semantic interoperability in such kind of networks. Ontologies are used for description of knowledge domains. A notation of object-oriented constraint networks is proposed for ontology modeling. Any-to-one decentralized model has been selected for semantic interoperability description and implemented using distribution of ontology slices between nodes of the network.

Keywords: decision making support, ontology management, self-organizing networks.

sai: cosiwn.2007.09.136

Economic Coordination in Kanban-Based Production Systems with Ubiquitous Computing

Titus Faupel and Stefan Sackmann

System and Information Sciences Notes, Vol. 2, No. 1, September 2007, pp. 33-39

Abstract: Ubiquitous computing has the potential to decrease coordination costs and thereby change the relative advantageousness of various coordination mechanisms. In this contribution, an economically aligned decentralized coordination mechanism for autonomous sequence planning in Kanban processes is developed. A first evaluation of the improvement of the coordination results is accomplished and discussed by means of a simulation study.

Keywords: Kanban, self-coordination, ubiquitous computing.

sai: cosiwn.2007.09.137

Stable Collaboration Patterns of Self-Interested Agents in Iterative Request for Proposal Coalition Formation Environments

Carlos Merida-Campos and Steven Willmott

System and Information Sciences Notes, Vol. 2, No. 1, September 2007, pp. 40-45

Abstract: This paper explores a general model of economic exchange between heterogeneous agents representing firms, traders, or other socioeconomic entities, that self-organise into coalitions to fulfil specific tasks. In particular, the work addresses coalition formation problems in which many tasks are addressed to the same population over time in an iterative fashion. The purpose of the paper is to describe the necessary elements that lead the system to an equilibrium state and asses the impact of coalition size constraints on the type of collaboration patterns established between agents. By using a novel data mining technique called collaboration graphs it is possible to see that stable states can be reached using simple iterative protocols and that the number of stable states increases as the coalition size limit decreases.

Keywords: MAS, Coalition formation, Request for Proposal, distributed problem solving, Electronic markets.

sai: cosiwn.2007.09.138

The Emergence of Order in Goods Distribution Using Information and Competition

David Cabanillas and Steven Willmott

System and Information Sciences Notes, Vol. 2, No. 1, September 2007, pp. 46-51

Abstract: This paper is concerned with the feasibility of achieving a competitive allocation of goods in a decentralized context. The distribution of a set of goods amongst a set of distributed and autonomous agents is a complex combinatorial problem. The three main challenges involved are: 1) the detection of needs by their members, 2) the structure of the environment within which exchanges occur, and 3) strategic users acting selfishly. A market–based mechanism provides a powerful approach to regulate exchanges between members of a community competing to maximize their utility/satisfaction and multiple factors such as topology and self-interest influence the decision making process. Therefore, we study the emergence of order with respect to goods distribution in a regulated market–based approach and how information and competition affect market performance.

Keywords: allocation, bartering, market–based approach.

sai: cosiwn.2007.09.139

Exploiting MAS Self-Organization for Distributed Constraint Satisfaction Problems

Giovanni Pezzulo

System and Information Sciences Notes, Vol. 2, No. 1, September 2007, pp. 52-57

Abstract: Multi Agent Systems are especially suited for solving problems that require collective efforts and cooperation. It is generally assumed that resources such as knowledge and capabilities are distributed among agents. In this paper we illustrate the functioning of a MAS framework, AKIRA, in which energetic resources (i.e., speed and memory) are also distributed among the agents according to two typical principles of self-organizing systems, local excitation and global inhibition. We then demonstrate its behavior in a distributed constraint satisfaction task, the 3-coloring problem, in which each agent has to optimize a local function (its color) and can partially influence the other agents (by exchanging messages). At the same time, the whole MAS has to optimize a global function –a feature that is shared by several real world problems, including traffic management, networking, etc. Our results show that self-organization of agents and resources boundedness of the MAS permit to give priority to the better tentative solutions, and at the same time to identify and exploit hidden relations among them, two of the most critical problems of distributed problem solving.

Keywords: DCSP, MAS, self-organization, agents.

sai:cosiwn.2007.09.140

Self-Adaptive Execution Mechanisms in Adaptive Resource Management System

Daniele Cammareri and Claudia Raibulet

System and Information Sciences Notes, Vol. 2, No. 1, September 2007, pp. 58-63

Abstract: This paper presents our approach to self-adaptation in the context of an Adaptive Resource Management System (ARMS). The main objectives of ARMS are (1) to identify the most appropriate system component able to execute the current service and (2) to ensure that the execution of services meets the qualities required and expected by users. To achieve the first objective, ARMS exploits reflection through which it is able to monitor the underlying system's components and the services they provide in order to choose the best solution for the current request. The second objective requires run-time adaptation mechanisms to monitor the actual execution of services and to perform changes whenever an execution cannot be fulfilled or it does not meet the requested qualities. In this paper, attention is focused on the solution designed and implemented to achieve this last objective. We present the adaptive and decision support mechanisms considered to observe and control the execution of services. The paper introduces also the main issues to be addressed when dealing with self-adaptive systems, issues which range from analysis to design and implementation.

Keywords: adaptive service execution, self-adaptation.

sai: cosiwn.2007.09.141

Beyond Selection: A Case for Fine-Grained Complex System Modelling

Mariusz Nowostawski

System and Information Sciences Notes, Vol. 2, No. 1, September 2007, pp. 64-69

Abstract: The motivation for this study is to investigate whether selection/fitness-based models are sufficient to address adequately the main challenges of modelling evolutionary dynamics, or rather should they be augmented/replaced with explicit hierarchical and fine-grained evolutionary models. We briefly discuss limitations of the selection/fitness models, then present modelling approaches used within the field of artificial life and artificial chemistry that do not use selection/fitness modelling directly. Finally, we present our own research framework based on computational meta-evolutionary approach, called Evolvable Virtual Machines (EVM). The objectives of this article are three-fold. First, we identify and review some of the long standing problems in evolutionary biology and artificial life. Second, we argue the need for integration, generalisation and a shift towards fine-grained, detailed, and non-probabilistic evolutionary modelling techniques to successfully address and solve these problems. Third, we present some of the currently used techniques for modelling evolutionary dynamics without the direct use of selection/fitness abstractions.

Keywords: causality, coarse-grained and fine-grained modelling, modelling evolutionary dynamics, selection-based models.

sai: cosiwn.2007.09.142

An Agent Based Simulation for Testing the Emergence of Meaning

Toomas Kirt

System and Information Sciences Notes, Vol. 2, No. 1, September 2007, pp. 70-73

Abstract: To understand the essence of meaning is a crucial point to build intelligent systems. It is proposed that meaning emerges if an agent starts to distinguish objects or events that have positive or negative impact on survival and to prefer desirable and avoid undesirable states. In this paper a simulation is proposed to evaluate whether it is possible that from a random initial configuration with the help of an evolutionary process an evaluation system emerges that helps an agent to distinguish and gather energy rich resources and to avoid dangerous matter.

Keywords: artificial life, evolutionary computation, meaning, agents.

sai: cosiwn.2007.09.143

Meta-Models and Grid-Services for Management of Clinical Practice Data for Evidence-Based Pharmacology

Alexey Zhuchkov, Nikolay Tverdokhlebov, Boris Alperovich and Alexander Kravchenko

System and Information Sciences Notes, Vol. 2, No. 1, September 2007, pp. 74-77

Abstract: The paper describes XML-based information structures (meta-models) and software services developed to search, collect and aggregate data of drug clinical usage and to represent the information to users in a comprehensible way. These tools provide biomedical professionals with complete, detailed and actual information on possible adverse events related to drug administration and thus support the evidence-based pharmacology and more personalized healthcare. Grid provides implementation of these tools on base of multi-agent system technology enabling information processing in loosely coupled data space.

Keywords: clinical data management, grid-service, XML meta-model.

sai: cosiwn.2007.09.144

An Optimal Approach To Determine The Minimum Architecture for Real-Time Embedded Systems Scheduled by EDF

Jean-François Hermant and Laurent George

System and Information Sciences Notes, Vol. 2, No. 1, September 2007, pp. 78-81

Abstract: This paper presents a sensitivity analysis on the Worst-Case Execution Times of sporadic tasks for the dimensioning of real-time embedded systems in which tasks are executed according to the preemptive Earliest Deadline First (EDF) scheduling policy. The timeliness constraints of the tasks are expressed in terms of late termination deadlines. A general case is considered, where the task deadlines are independent of the task sporadicity intervals (also called periods). New results for EDF are shown, which enable us to determine the minimum architecture corresponding to the minimum processing speed architecture such that all the task deadlines are met. This minimum architecture is obtained from the analysis of EDF in a reference architecture in a time interval of length equals to the least common multiple of the task periods. From this analysis, it is then straightforward to determine, if the sporadic task set is feasible with another processor speed.

Keywords: Real-Time Scheduling, Embedded Systems, Earliest Deadline First, Sensitivity Analysis, C-space, Feasibility Domain, Minimum Architecture.

sai: cosiwn.2007.09.145

Towards Collaborative Coping with IT Complexity by Combining Service-Oriented Architectures and Organic Computing

Stefan Thanheiser, Lei Liu and Hartmut Schmeck

System and Information Sciences Notes, Vol. 2, No. 1, September 2007, pp. 82-87

Abstract: By applying the paradigm of service-oriented architectures (SOA) to existing IT infrastructures, enterprises try to achieve complexity reduction concerning their business IT. However, latest research indicates the presence of an SOA-specific complexity, whose impact on total IT complexity is hard to measure. A plausible strategy for better coping with IT complexity lies in introducing self-organising IT systems as proposed by the ‘Organic Computing’ Initiative. To discuss how a relative reduction of IT complexity to human system participants can be accomplished, this paper presents the approach of an ‘organic SOA’ aiming at collaborative handling of complexity by means of introducing self-organizing systems - and yet maintaining SOA flexibility.

Keywords: controlled self-organization, IT complexity, organic computing, service-oriented architectures.

sai: cosiwn.2007.09.146

Context-Aware Adaptation for Group Communication Support Applications with Dynamic Architecture

Ismael Bouassida Rodriguez, Khalil Drira, Christophe Chassot and Mohamed Jmaiel

System and Information Sciences Notes, Vol. 2, No. 1, September 2007, pp. 88-92

Abstract: In this paper, we propose a refinement-based adaptation approach for the architecture of distributed group communication support applications. Unlike most of previous works, our approach reaches implementable, context-aware and dynamically adaptable architectures. To model the context, we manage simultaneously four parameters that influence Qos provided by the application. These parameters are: the available bandwidth, the exchanged data communication priority, the energy level and the available memory for processing. These parameters make it possible to refine the choice between the various architectural configurations when passing from a given abstraction level to the lower level which implements it. Our approach allows the importance degree associated with each parameter to be adapted dynamically. To implement adaptation, we switch between the various configurations of the same level, and we modify the state of the entities of a given configuration when necessary. We adopt the direct and mediated Producer-Consumer architectural styles and graphs for architecture modelling. In order to validate our approach we elaborate a simulation model.

Keywords: adaptation, context-aware, graphs, producer-consumer style, software architecture.

sai: cosiwn.2007.09.147

Self-Organization and Adaptation in Socio-Cognitive Systems: A Computational Model

Christina Klüver and Jürgen Klüver

System and Information Sciences Notes, Vol. 2, No. 1, September 2007, pp. 93-96

Abstract: The concept of adaptation means either that a whole system is adaptive or that the elements of the system are so and generate that way the adaptability of the system. The modeling of socio-cognitive requires the second form of adaptation. We describe this form of adaptation by a model of the evolution of social systems via the individual learning processes of the elements of the system, i.e. the social actors. Each artificial actor consists of different neural networks.

Keywords: cognitive learning processes, socio-cognitive systems, social self-organization.

sai: cosiwn.2007.09.148

Reactive Common Sense Reasoning for Knowledge-based Self-Optimization

Michael Cebulla

System and Information Sciences Notes, Vol. 2, No. 1, September 2007, pp. 97-100

Abstract: We discuss a membrane-based calculus for the combination of conceptual spaces during runtime. We claim that properties like self-optimization and context-adaptive behavior can be supported by the runtime combination of such situational models. Since our goal is to support emergent properties of behavior (and due to the fact that it is not possible to define a complete calculus for all situations) we introduce terms which are capable of self-modification. Terms from situational descriptions can evolve according to simple rules thus providing various possibilities for reactions. This strategy is well-suited to support a decentral approach towards the modeling of distributed behavior.

Keywords: Self-Optimization, Common Sense- Reasoning, Situation Awareness.

sai: cosiwn.2007.09.149

Software Engineering Concerns in Observing Networks of Autonomic Systems

David Lamb, Martin Randles and A. Taleb-Bendiab

System and Information Sciences Notes, Vol. 2, No. 1, September 2007, pp. 101-104

Abstract: To address the lack of scalability in current models of autonomic system design, based on centralised feedback control, much research is underway to investigate alternative decentralised self-organising system approaches. However, their engineering support is limited, and not well understood. As such, this paper, describing ongoing research into “engineering emergence”, presents results related to signature-based methods for detection and characterisation of emerging structure in networks of autonomic systems. The approach provides a specification of observed structure, plus that of a matching and efficient observation overlay. It is tested via a simulation environment to demonstrate how observers can make efficient use of limited resources while accurately observing a system.

Keywords: autonomic systems, observation, overlay, self-organising networks, software engineering.

sai: cosiwn.2007.09.150

A Simulation Study of Grid Scheduling

Petros Papadopoulos, Huaglory Tianfield and Mike Mannion

System and Information Sciences Notes, Vol. 2, No. 1, September 2007, pp. 105-109

Abstract: This paper conducts an initial study of existing Grid scheduling solutions and performs simulations to test common Grid scheduling algorithms.

Keywords: grid, grid scheduling, grid simulation

sai:cosiwn.2007.09.151

An Agent Based Component for QoS Management in a Grid-P2P Architecture

Antonella Di Stefano, Giovanni Morana and Daniele Zito

System and Information Sciences Notes, Vol. 2, No. 1, September 2007, pp. 110-113

Abstract: Many of the middlewares capable of supporting Grid computing envisage trust among participants, but they often adopt centralized approaches that present problems related to scalability and fault tolerance. On the other hand, P2P systems have decentralized models that are intrinsically scalable and fault tolerant than the Grid systems, even if, due to security-related and credential management issues, they represent a poor solution.

This paper proposes an architecture that integrates the P2P interaction model in Grid environments, so as to build an open cooperative model wherein Grid entities, belonging to different administrative domains, are composed in a decentralized way. We focus our attention on AGENDA, a module within the architecture that allows decentralization by spreading the information system over all the nodes of the Grid. In particular, AGENDA handles Service Level Agreement and supplies a flexible interface enabling the client to negotiate QoS requirement with resource providers, exploiting the low level advanced reservation functionalities.

Keywords: P2P Grid, agents system, QoS.

sai:cosiwn.2007.09.152

Engineering Self-Management into Legacy Systems

Jens Steiner and Ursula Goltz

System and Information Sciences Notes, Vol. 2, No. 1, September 2007, pp. 114-117

Abstract: For a few years now, the ever increasing complexity of technical systems has been one of the major obstacles for further advancements. Inspired by nature, concepts like self-management and self-organization have found their way into artificial systems which in turn exhibit self-optimization, self-healing or other so called self-* properties. While first engineering approaches for such systems exist, there is no methodology yet, that is capable to infuse self-management into legacy systems and in addition can prove that functional and non functional requirements are still met after the reengineering. This paper proposes a methodical approach for this task, emphasizing the use of model-based lightweight and formal methods for validation and verification.

Keywords: autonomic computing, methodology, model-based development, self-management, self-* properties, validation, verification.

sai: cosiwn.2007.09.153

Flexible Application and Context Aware Adaptation in a Pervasive File System

Gustavo C. Frainer, Luciano da Silva, Iara Augustin, Adenauer Yamin and Cláudio Geyer

System and Information Sciences Notes, Vol. 2, No. 1, September 2007, pp. 118-121

Abstract: This paper presents the Pervasive File Space (PFS), a service that provides pervasive access to files, using context and application aware adaptation to provide a better service without burdening the user. The PFS introduces a new model for application-aware adaptation that enables the it to have a larger number of adaptative behaviors and to deal with any context element the application deems important.

Keywords: application-aware adaptation, file system, pervasive computing.

sai:cosiwn.2007.09.154

Protein Secondary Structure Prediction through a Cooperative Multiagent Learning Approach

Andrea Addis, Giuliano Armano, Francesco Mascia and Eloisa Vargiu

System and Information Sciences Notes, Vol. 2, No. 1, September 2007, pp. 122-125

Abstract: This paper illustrates a cooperative multiagent learning approach devised to perform classification or prediction tasks. The resulting system is composed by a population of agents that cooperate and interact in accordance with generic requirements imposed by the adoption of evolutionary computation strategies. As a case study, we consider the typical bioinformatics problem of predicting protein secondary structure.

Keywords: evolutionary computation, multiagent systems, soft computing

sai: cosiwn.2007.09.155

Market-Based Coordination Strategies for Large-scale Multi-Agent Systems

MyungJoo Ham and Gul Agha

System and Information Sciences Notes, Vol. 2, No. 1, September 2007, pp. 126-131

Abstract: This paper studies market-based mechanisms for dynamic coordinated task assignment in large scale agent systems carrying out search and rescue missions. Specifically, the effect of different auction mechanisms and swapping are studied. The paper describes results from a large number of simulations of homogeneous agents, where by homogeneous we mean that agents in a given simulation use the same strategy. The information available to agents and their bidding strategies are used as simulation parameters. The simulations provide insight about the interaction between the strategy used by individual agents and the market mechanism. Performance is evaluated using several metrics: mission time, distance traveled, communication and computation costs, and workload distribution. Some of the results obtained include: limiting information may improve performance, different utility functions may affect the performance in non-uniform ways, and swapping may help improve the efficiency of assignments in dynamic environments.

Keywords: auction, market-based approach, multi-agent system, task assignment.

sai:cosiwn.2007.09.156

ASSL Specification of a Self-Scheduling Mechanism in Team-Robotics Modeled with the AS-TRM

Olga Ormandjieva and Emil Vassev

System and Information Sciences Notes, Vol. 2, No. 1, September 2007, pp. 132-137

Abstract: The objective of this research is a rigorous investigation of the Autonomic Systems Timed Reactive Model (AS-TRM) architectural approach for developing and evolving reactive autonomic (self-managing) team-robotics systems, and for specifying the continuous self-scheduling of team tasks and the self-monitoring of task fulfillment. The Autonomic System Specification Language (ASSL) is a framework for formally specifying and generating autonomic systems. With ASSL, we can specify high-level behavior policies, thus it is a highly appropriate language for specifying a self-scheduling mechanism as part of the overall team robotics system behavior. An appropriate architecture for supporting self-scheduling for a supervised team of mobile, autonomic and reactive robots, along with a communication mechanism to implement the behavior of these robotic systems, are specified with ASSL.

Keywords: autonomic system, formal specification language, reactive system, self-scheduling, team robotics.

----------------------------------------------------------------------------------------------

Copyright © 2007 Systemics and Informatics World Network