Approximating the empirical distribution of a measured data trace by a phase-type (PH) distribution has significant applications in the analysis of stochastic models. For phase-type fitting, a large number of different methods and tools exist. One drawback of all these methods is that the fitting effort strongly depends on the size of the data trace to be fitted. Since large data traces are necessary to capture rare events, which have a strong impact on system performance, current fitting procedures are very time consuming. In this paper, we introduce a method to generate an aggregated trace from the original trace, and we show how to effectively use the aggregated trace within a PH fitting approach, called G-FIT. In particular, we show that elements of a large traffic trace can be aggregated to a smaller number of 50–200 weighted elements, while fitting accuracy remains the same compared to the case of fitting the original trace. As a result, CPU time requirements for fitting PH distributions can be decreased by about four orders of magnitude, such that traces with ten million elements can be accurately fitted in a few seconds. The effectiveness of the proposed method is demonstrated on a set of benchmark traces and two real traffic traces as well as quantitative results from queuing analysis.

More and more complex software systems in modern vehicles require development processes being based on explicit software and system architecture models. If used in early phases of the process, errors and hazards can be identified when modelling new aspects of the overall system. This article will use the dSpace’s tool chain, including the company’s latest product SystemDesk, to outline possible implementations of these ideas.

Immer komplexere Softwaresysteme in modernen Fahrzeugen erfordern Entwicklungsprozesse, die auf expliziten Software- und Systemarchitekturmodellen basieren. Werden solche Modelle schon in frühen Phasen des Prozesses eingesetzt, können Fehler und Gefahrenquellen bereits bei der Modellierung neuer Aspekte des Gesamtsystems erkannt werden. In diesem Beitrag wird anhand der Werkzeugkette von dSpace einschließlich des neuesten Produkts SystemDesk erläutert, wie sich diese Methoden umsetzen lassen.

In general, decision support is one of the main purposes of model-based analysis of systems. Response surface methodology (RSM) is an optimization technique that has been applied frequently in practice, but few automated variants are currently available. In this paper, we show how to combine RSM with numerical analysis methods to optimize continuous time Markov chain models. Among the many known numerical solution methods for large Markov chains, we consider a Gauss-Seidel solver with relaxation that relies on a hierarchical Kronecker representation as implemented in the APNN Toolbox. To effectively apply RSM for optimizing numerical models, we propose three strategies which are shown to reduce the required number of iterations of the numerical solver. With a set of experiments, we evaluate the proposed strategies with a model of a production line and apply them to optimize a class-based queueing system.

A model-based design of systems requires appropriate tool support in many ways. It requires a modeling notation that suits the application problem, a set of analysis techniques that provide qualitative and/or quantitative results, and finally some optimization methods that help a designer to make appropriate design decisions. The challenge is to integrate those components into a homogenous framework such that a model based design takes advantage from synergy effects that result from a sophisticated combination of modeling formalism, analysis and optimization technique. In this paper, we present OPEDo, a tool framework that integrates modeling tools and analysis engines with state-of-the-art optimization methods. With respect to modeling, it contains the ProC/B editor for specifying open process-oriented simulation models, the APNN Toolbox for modeling with stochastic Petri nets, and OMNet++, for modeling using a simulation language. OPEDo provides analysis techniques for stochastic models based on discrete event simulation, based on queueing network analysis and numerical analysis techniques for continuous time Markov chains with the help of HIT, OMNeT++, and APNN Toolbox. Optimization of stochastic models has particular challenges due to the cost of model evaluation and the precision of results that can be achieved, so OPEDo contains specially adjusted variants of a variety of optimization methods, which includes response surface methodology, evolutionary strategies, genetic algorithms, and Kriging metamodeling techniques.

The representation of general distributions or measured data by phase-type distributions is an important and nontrivial task in analytical modeling. Although a large number of different methods for fitting parameters of phase-type distributions to data traces exist, many approaches lack efficiency and numerical stability. In this paper, a novel approach is presented that fits a restricted class of phase-type distributions, namely, mixtures of Erlang distributions, to trace data. For the parameter fitting, an algorithm of the expectation maximization type is developed. This paper shows that these choices result in a very efficient and numerically stable approach which yields phase-type approximations for a wide range of data traces that are as good or better than approximations computed with other less efficient and less stable fitting methods. To illustrate the effectiveness of the proposed fitting algorithm, we present comparative results for our approach and two other methods using six benchmark traces and two real traffic traces as well as quantitative results from queuing analysis

In most cases, model based analysis of discrete event systems is part of a more general process with the ultimate goal to find an optimal configuration for the modeled system. Consequently, there is a need for a seamless integration of analysis techniques and optimization methods into a framework in order to provide adequate support for the design of systems. However, most tools for performance or dependability modeling do not include support for optimization. In addition, most optimization methods require substantial refinements to match the challenging properties of stochastic models, e.g., that a model evaluation may yield statistical estimates instead of precise results, that a model evaluation may be computationally expensive, and that a model evaluation does not provide derivatives of the objective function. Currently optimization of discrete event systems is often done in an ad hoc manner by hand. The goal of OPEDo is twofold: First to customize state-of-the-art optimization methods for the specific needs of discrete event systems and second to combine these optimization methods with state-of-the-art approaches for the specification and analysis of discrete event systems.

In this paper, we present an evolution strategy for the optimization of simulation models. Our approach incorporates statistical selection procedures that efficiently select the best individual, where best is defined by the maximum or minimum expected simulation response. We use statistical procedures for the survivor selection during the evolutionary process and for selecting the best individual from a set of candidate best individuals, a so-called elite population, at the end of the evolutionary process. Furthermore, we propose a heuristic selection procedure that reduces a random-size subset, containing the best individual, to at most a predefined size. By means of a stochastic sphere function and a simulation model of a production line, we show that this procedure performs better in terms of number of model evaluations and solution quality than other state-of-the-art statistical selection procedures.

The representation of general distributions or measured data by phase-type distributions is an important and non-trivial task in analytical modeling. Although a large number of different methods for fitting parameters of phase-type distributions to data traces exist, many approaches lack efficiency and numerical stability. In this paper, a novel approach is presented that fits a restricted class of phase-type distributions, namely mixtures of Erlang distributions, to trace data. For the parameter fitting an algorithm of the expectation maximization type is developed The paper shows that these choices result in a very efficient and numerically stable approach which yields phase-type approximations for a wide range of data traces that are as good or better than approximations computed with other less efficient and less stable fitting methods. To illustrate the effectiveness of the proposed fitting algorithm, we present comparative results for our approach and two other methods using six benchmark traces and two real traffic traces.

Large logistics networks often require sophisticated decisions to be made to meet the required service qualities. Often these decisions are made according to a model based analysis and optimization of the network. For this purpose simulation models and appropriate optimization techniques have to be combined. This combination is still a challenge, in particular if the approach should run in a more or less automated way. In this paper we present the combination of a process chain based simulator and the response surface method for optimization. Particular emphasis is placed onto a realization of the response surface method which runs completely automatically after initialization. The quality of the proposed optimization approach is shown by means of two example models.

In general, decision support is one of the main purposes of model-based analysis of systems. Response surface methodology (RSM) is an optimization technique that has been applied frequently in practice, but few automated variants are currently available. In this paper, we propose the combination of RSM with numerical analysis methods to solve continuous time Markov chain models of class-based queueing systems (CBQ). We consider first- and second-order models in RSM to identify an optimal parameter configuration for CBQ as part of the differentiated service architecture. Among the many known numerical solution methods for large Markov chains, we consider a Gauss-Seidel solver with relaxation that relies on a hierarchical Kronecker representation as implemented in the APNN Toolbox. To effectively apply the proposed optimization methodology we determine a suitable configuration of RSM and compare the results with previous results for optimizing CBQ.

In this paper, we show how online management of both quality of service (QoS) and provider revenue can be performed in CDMA cellular networks by adaptive control of system parameters to changing traffic conditions. The key contribution is the introduction of a novel call admission control and bandwidth degradation scheme for real-time traffic as well as the development of a Markov model for the admission controller. This Markov model incorporates important features of 3G cellular networks, such as CDMA intra- and inter-cell interference, different call priorities and soft handover. From the results of the Markov model the threshold for maximal call degradation is periodically adjusted according to the currently measured traffic in the radio access network. As a consequence, QoS and revenue measures can be optimized with respect to a predefined goal. To illustrate the effectiveness of the proposed QoS/revenue management approach, we present quantitative results for the Markov model and a comprehensive simulation study considering a half-day window of a daily usage pattern.

Third-generation (3G) mobile communication systems are currently one of the key communication technologies in research and development due to the high market demand for advanced wireless communication. The current evolution is primarily characterized by a transition from circuit-switched voice-oriented networks to integrated multi-service all IP networks. To effectively design complex mobile communication systems, the design process should be accompanied by stochastic modeling and quantitative evaluation of different design alternatives. The most popular language for model specification used in industrial projects is the Unified Modeling Language (UML). Although conceived as a general-purpose modeling language, the current version of the UML does not contain building blocks for introducing stochastic timing into UML diagrams. The first part of this thesis presents new results for numerical quantitative analysis of discreteevent stochastic systems specified in Petri net notation or as UML diagrams. An efficient algorithm for the state space generation out of an UML state diagram or activity diagram that allows quantitative analysis by means of the underlying stochastic process is presented. Furthermore, this thesis considers new methodological results for the effective numerical analysis of finite-state generalized semi-Markov processes with exponential and deterministic events by an embedded general state space Markov chain (GSSMC). Key contributions constitute (i) the observation that elements of the transition kernel of the GSSMC can always be computed by appropriate summation of transient state probabilities of continuous-time Markov chains and (ii) the derivation of conditions under which kernel elements are constant. To provide automated tool support, the presented algorithms are included in the software package DSPNexpress-NG available for download on the Web. The support of multimedia services over wireless channels presents a number of technical challenges. One of the major challenges is to effectively utilize the scarce radio bandwidth in the access network by adaptive control of system parameters. The second part of this thesis is devoted to this topic. A Markov model representing the sharing of radio channels by circuitswitched connections and packet-switched sessions under a dynamic channel allocation scheme is evaluated. Closing the loop between network operation and network control, a framework for the adaptive quality of service management for 3G mobile networks is introduced. Building on this framework, a novel call admission control and bandwidth reservation scheme for the optimization of quality of service for mobile subscribers is presented. The performance of the solutions proposed in this thesis is investigated experimentally based on numerical quantitative analysis and discrete-event simulation.

In this paper, we introduce a unified approach for the adaptive control of 3G mobile networks in order to improve both quality of service (QoS) for mobile subscribers and to increase revenue for service providers. The introduced approach constantly monitors QoS measures as packet loss probability and the current number of active mobile users during operation of the network. Based on the values of the QoS measures just observed, the system parameters of the admission controller and packet scheduler are controlled by the adaptive performance management entity. Considering UMTS, we present performance curves showing that handover failure probability is improved by more than one order of magnitude. Moreover, the packet loss probability can be effectively regulated to a predefined level and provider revenue is significantly increased for all pricing policies.

In this paper, we introduce a framework for the adaptive control of universal mobile telecommunications system (UMTS) networks in order to improve bandwidth utilization of the radio channels. The key contribution of the paper constitutes the introduction of a performance management information base for dynamically adjusting the packet scheduler and admission controller. Thus, the adaptive control framework closes the loop between network operation and network control. Furthermore, the adaptive control framework can effectively deal with the different time scales of packet scheduling and admission control. Moreover, we present a traffic model for non-real-time UMTS traffic based on measured trace data. The analysis and scaling process of the measured trace data with respect to different UMTS bandwidth classes constitutes the basic concept of this traffic characterization. Using this traffic model and simulation on the IP level, the gain of employing the adaptive control framework is illustrated by performance curves for various quality of service measures.

In this paper, we propose extensions to UML state diagrams and activity diagrams in order to allow the association of events with exponentially distributed and deterministic delays. We present an efficient algorithm for the state space generation out of these UML diagrams that allows a quantitative analysis by means of an underlying stochastic process. We identify a particular stochastic process, the generalized semi-Markov process (GSMP), as the appropriate vehicle on which quantitative analysis is performed. As innovative feature the algorithm removes vanishing states, i.e. states with no timed events active, and considers branching probabilities within activity diagrams. Furthermore, we introduce a performance evaluation framework that allows a system designer to predict performance measures at several steps in the design process. The applicability of our approach for practical performance and dependability projects is demonstrated by an UML specification of the General Packet Radio Service, a packet switched extension in GSM wireless networks.

This paper presents a discrete-event simulator for the General Packet Radio Service (GPRS) on the IP level. GPRS is a standard on packet data in GSM systems that will become commercially available by the end of this year. The simulator focuses on the communication over the radio interface, because it is one of the central aspects of GPRS. We study the correlation of GSM and GPRS users by a static and dynamic channel allocation scheme. In contrast to previous work, our approach represents the mobility of users through arrival rates of new GSM and GPRS users as well as handover rates of GSM and GPRS users from neighboring cells. Furthermore, we consider users with different QoS profiles modeled by a weighted fair queueing scheme. The simulator considers a cell cluster comprising seven hexagonal cells. We provide curves for average carried traffic and packet loss probabilities for different channel allocation schemes and packet priorities as well as curves for average throughput per GPRS user. A detailed comparison between static and dynamic channel allocation schemes is provided.

Presents an efficient and accurate analytical model for the radio interface of the General Packet Radio Service (GPRS) in a GSM network. The model is utilized for investigating how many packet data channels should be allocated for GPRS under a given amount of traffic in order to guarantee appropriate quality of service. The presented model constitutes a continuous-time Markov chain. The Markov model represents the sharing of radio channels by circuit-switched GSM connections and packet-switched GPRS sessions under a dynamic channel allocation scheme. In contrast to previous work, the Markov model explicitly represents the mobility of users by taking into account arrivals of new GSM and GPRS users as well as handovers from neighboring cells. To validate the simplifications necessary for making the Markov model amenable to numerical solution, we provide a comparison of the results of the Markov model with a detailed simulator on the IP level.

Article Free Access Share on Quantitative system evaluation with DSPNexpress 2000 Authors: Christoph Lindemann University of Dortmund, Department of Computer Science, August-Schmidt-Str. 12, 44227 Dortmund, Germany University of Dortmund, Department of Computer Science, August-Schmidt-Str. 12, 44227 Dortmund, GermanyView Profile , Axel Thümmler University of Dortmund, Department of Computer Science, August-Schmidt-Str. 12, 44227 Dortmund, Germany University of Dortmund, Department of Computer Science, August-Schmidt-Str. 12, 44227 Dortmund, GermanyView Profile , Alexander Klemm University of Dortmund, Department of Computer Science, August-Schmidt-Str. 12, 44227 Dortmund, Germany University of Dortmund, Department of Computer Science, August-Schmidt-Str. 12, 44227 Dortmund, GermanyView Profile , Marco Lohmann University of Dortmund, Department of Computer Science, August-Schmidt-Str. 12, 44227 Dortmund, Germany University of Dortmund, Department of Computer Science, August-Schmidt-Str. 12, 44227 Dortmund, GermanyView Profile , Oliver P. Waldhorst University of Dortmund, Department of Computer Science, August-Schmidt-Str. 12, 44227 Dortmund, Germany University of Dortmund, Department of Computer Science, August-Schmidt-Str. 12, 44227 Dortmund, GermanyView Profile Authors Info & Claims WOSP '00: Proceedings of the 2nd international workshop on Software and performanceSeptember 2000 Pages 12–17https://doi.org/10.1145/350391.350396Online:01 September 2000Publication History 7citation345DownloadsMetricsTotal Citations7Total Downloads345Last 12 Months3Last 6 weeks1 Get Citation AlertsNew Citation Alert added!This alert has been successfully added and will be sent to:You will be notified whenever a record that you have chosen has been cited.To manage your alert preferences, click on the button below.Manage my Alerts New Citation Alert!Please log in to your account Save to BinderSave to BinderCreate a New BinderNameCancelCreateExport CitationPublisher SiteeReaderPDF