I will describe the principles involved in the design of software architecture. The premise is that the design of software architecture is determined by the quality attributes requirements for a system. The quality attributes that we have focused on are: availability, modifiability, performance, and usability.

Our premise leads us to
1. specifying the quality attribute requirements in a common form
2. cataloging design primitives that support the achievement of each quality attribute, and
3. having mechanisms for managing trade offs among the attributes.

A software architecture for a system is the structure or structures of the system, which comprise elements, the externally visible properties of those elements, and the relationships between them. The software architecture for a system largely determines the system's quality attributes such as performance or modifiability, and it serves to constrain and guide the implementation. The software architecture serves as the vehicle for negotiating tradeoffs among requirements and stakeholder needs, and for tracking progress of teams. In order for the architecture to serve all these important roles, it must be effectively communicated -- that is, written down -- so that everyone who uses it can understand it. Hence, architecture documentation has emerged as a critical part of the architecting process. Modern treatments of architecture take a view-based approach, and in this session we will present a view-based documentation approach developed at the Software Engineering Institute for capturing software architectures. We will cover a set of very useful views and show how to capture each one using UML and other notations. Finally we will discuss a procedure for choosing which views of an architecture to capture in its documentation.

Software has been evolving from pre-defined, monolithic, centralized architectures to increasingly decentralized, distributed, dynamically composed federations of components. Binding mechanisms have been evolving from static to highly dynamic schemes based on discovery, negotiation and optimization. The talk surveys the evolution of software composition mechanisms, from "in-the-small" embedded environments to "in-the-large" environments like web services.

The wide acceptance of the Java programming language with its built-in concurrency constructs means that concurrent programming is no longer restricted to the minority of programmers involved in operating systems and embedded real-time applications. Concurrency is useful in a wide range of applications. Errors in these applications and systems may be life threatening, adversely affect our quality of life or may have severe financial implications. An understanding of the principles of concurrent programming and an appreciation of how it is practiced is essential for software engineering professionals. This tutorial takes a model-based approach to the design of concurrent and distributed programs. The models, based on finite state machines, provide insight into the composite behaviour of the programs, and can be automatically analysed to check various properties before implementation. The approach, concurrency concepts and analysis will be demonstrated through a series of examples and using the LTSA toolkit. In addition, current research work in the area will be discussed and demonstrated.

The Eiffel method promotes the construction of quality software through a number of principles such as seamless development, to ensure the continuity of the software process, and Design by Contract. It takes advantage of object-oriented techniques including multiple inheritance and genericity, and covers the entire lifecycle. This set of lectures presents an in-depth introduction to the method and its applications, sufficient to start building quality reusable components and perform contract-based analysis and design.

Hardware-software codesign is a recent research area whose biggest focus is to facilitate the design of small embedded systems, that is, those consisting of a microprocessor (such as DSP), for which software is required, plus one or more semicustom ASICs or FPGAs (field programmable gate arrays). Codesign methodologies are intended to give relief to designers struggling with ad hoc divisions of hardware and software components, and the attendant integration problems. Their purpose is to streamline the design process, thus reducing design costs and shortening time-to-market; to optimize the hardware/software partitioning, thus reducing direct product costs; and to ease integration, by automatically generating hardware/software interfaces.