Andreas Mitschele-Thiel

Performance engineering of SDL/MSC systems

The paper addresses issues related to the performance of systems specified with SDL and MSC and describes how performance aspects can be integrated with the SDL/MSC method. We identify the aspects that are relevant to performance evaluation and performance tuning. Afterwards we introduce basic performance modelling techniques and survey approaches that integrate performance evaluation into the context of SDL and MSC.

An Annotational Extension of Message Sequence Charts to Support Performance Engineering

Publisher Summary This chapter describes the performance message sequence charts (PMSCs), an extension of message sequence chart (MSC)-96 to include performance aspects into the MSC language. The language extension allows for the cospecification of functional and performance aspects. This joint specification of functional and performance aspects eases the task of software engineers. PMSC supports the formal specification of a variety of aspects related to performance. This includes performance requirements, resource requirements, the specification of the available resources as well as the specification of the mapping of the MSC instances on the available resources. To support the use of tools based on MSC-96, the language extensions are embedded in comments of the original MSC-96 language. The cospecification based on MSC rather than on Specification and Description Language (SDL) supports the early evaluation of the performance of the system.

Derivation of Efficient Implementations from SDL Specifications Employing Data Referencing, Integrated Packet Framing and Activity Threads*

Publisher Summary This chapter describes how three different techniques from the manual implementation of protocol architectures can be applied to automatically derive efficient implementations from Specification and Description Language (SDL) specifications. These three techniques are data referencing, integrated packet framing, and activity threads. Data referencing is employed to optimize communication among process instances. Integrated packet framing allows minimizing copying overhead within process instances. The activity thread model supports minimization of overhead for process management. The chapter analyzes the prerequisites to apply different optimization techniques and shows how the techniques can be employed for the automatic derivation of efficient implementations from SDL specifications. Finally, measurements comparing the performance achievable with the different optimization techniques are presented in the chapter. The measurements are made for different implementations of the Xpress Transfer Protocol derived from an SDL specification. The results obtained show that a considerable speedup can be achieved with each of the optimization techniques.

Design and optimization of high-performance protocols with the DO-IT toolbox

In the telecommunication industry, the Specification and Description Language (SDL) is a widely accepted technique to support the software development process. While several commercial SDL tools exist that focus on functional aspects, rather little research has been done concerning the integration of nonfunctional aspects in the development process.

On the Influence of Semantic Constraints on the Code Generation from Estelle Specifications

Implementations automatically derived from formal descriptions often do not fulfill the performance requirements of real-life applications. There are several reasons for this. In the paper, we discuss for the FDT Estelle how semantical constraints can influence the efficiency of the generated code. In the first part of the paper we show that certain language features may have a restraining effect on the performance of the implementation. The second part of the paper investigates how the activity thread model, a technique known from manual protocol implementation, can be applied to automatically derive efficient implementations from Estelle specifications. The activity thread model reduces the communication overhead. It is known to be efficient as the server model usually applied. We analyze the prerequisites to apply the model and present measurements comparing the performance achievable with the technique. The measurements are given for different implementations of XTP and the XDT protocol.

Methodology and tools for the development of high performance parallel systems with SDL/MSCs

The Specification and Description Language SDL and Message Sequence Charts (MSCs) are widely used in the telecommunication industry to support the software development process. In the paper, a methodology and a set of tools are described for the development of high performance parallel systems in the context of SDL and MSCs. While SDL and MSCs only support the formal specification of functional aspects of the system, we propose (1) the extension of MSCs to include non-functional requirements as the performance requirements of the application and (2) the annotation of SDL specifications with the respective execution cost on the parallel system. The formalization of non-functional aspects yields a set of benefits for system development: it allows the full integration of performance issues in all phases of the design process, starting from the requirements specification down to the final parallel implementation. It supports the automatization of performance related design decisions and allows the use of sophisticated tools supporting the performance optimization process.

Applying techniques and tools for the performance engineering of SDL systems

Abstract The integration of performance evaluation techniques into the SDL method is an important task. In particular in the area of telecommunications and distributed systems where product families have to be maintained for many years, performance should be considered from the very beginning of system design. This paper demonstrates how performance engineering can be successfully applied. Firstly, an SDL specification of TCP/IP is used to derive a performance model that includes the original SDL system as well as additional constructs describing the consumption of time and resources. Different implementation variants have been evaluated in experimental scenarios. Secondly, the application of a specification-driven monitoring technique to the same SDL specification of TCP/IP is described. The case study shows how the functional and temporal behaviour of an implementation derived from an SDL specification can be analysed and improved by monitoring the running system. The two approaches, modelling and monitoring, are supplementary to each other. This paper provides arguments that the systematic application of specification-driven performance engineering will result in major savings of time and money in later development phases and for later system releases.

Integrating model-based optimization and program transformation to generate efficient parallel programs

The paper gives an overview on the DSPL programming environment, an integrated approach to automate system design and implementation of applications run on dedicated parallel systems. The programming environment consists of a data-flow language and an integrated set of tools. The tools automatically derive a software model from the given application program. Based on the model, the design decisions as the network topology, the task mapping and schedule as well as the optimal use of buffers are computed. Finally, the design decisions are automatically implemented by transforming the application program in executable code for the chosen processor network. The DSPL programming environment integrates model-based optimization techniques and program transformation techniques. The integration allows to include new aspects in the optimization process. Especially optimizations crucial to the semantics of the program can be included. The most important examples of such optimizations are the enforcement of the schedule in case of data-dependent execution of tasks and the transformation of buffered communication to unbuffered communication. Both aspects are crucial to the generation of efficient parallel implementations. The integration of the two aspects is supported by a formal framework. This allows to formally prove the correctness of the program optimizations performed by the programming environment.

CORSAIR: HW/SW-Codesign von Kommunikationssystemen mit SDL

ZUSAMMENFASSUNG Der Artikel beschreibt die CORSAIR-Umgebung (Codesign and Rapid Prototyping System for Applications with Realtime Constraints) zum Entwurf und zur Implementierung von Kommunikationssystemen auf der Basis der formalen Beschreibungstechnik SDL. Im Unterschied zu gängigen SDL-Werkzeugen zielt das Projekt auf die Entwicklung gemischter HW/SW-Systeme anstatt reiner Softwareimplementierungen. Weitere Besonderheit unseres Ansatzes ist die durchgängige Berücksichtigung von Zeitanforderungen in allen Phasen des Entwicklungsprozesses. Unser Ansatz basiert auf einer Erweiterung von SDL, die die Beschreibung von Zeit- und Implementierungsaspekten unterstützt. CORSAIR zielt auf die weitgehende Automatisierung des Entwicklungsprozesses, entsprechend dem HW/SW-Codesign-Prinzip. HW/SW-Codesign unterstützt die späte und damit flexible Partitionierung der Spezifikation in Teile, die in Hardware bzw. in Software implementiert werden.

Die Realisierung der Manufacturing Message Specification mit der DSPL-Entwicklungsumgebung

Der Artikel beschreibt die Realisierung des Manufacturing Message Specification (MMS) Protokolls — dem wichtigsten Protokoll der Anwendungsschicht der Manufacturing Automation Protocol Specification (MAP) — auf einem Transputernetzwerk. Die Realisierung von MMS erfolgt mit der DSPL-Entwicklungsumgebung, einem Ansatz zur Unterstutzung der Abbildung und Optimierung von Anwendungen auf dedizierte parallele Systeme.