Alvin P. Mullery

Service architectures and service creation for integrated broadband communications

RACE is a collaborative, pre-competitive R&D programme, partially funded by the European Union. The programme comprises a number of projects involving participants from all major European telecommunications sectors. The focus is IBC, a concept which encompasses broadband fixed and mobile networks, broadcasting and multimedia communication services. The goal is to further the development and deployment of IBC in Europe, eventually forming a complete information and communication system available to, and usable by, all. RACE projects are grouped in Project Lines. Relevant to this paper is an area called Intelligence in Services and Networks (IS&N), which deals with the technologies required to engineer the next generation of the intelligent network or long term IN. Within the IS&N area, a number of projects have dealt with service engineering. This paper will present highlights of the vision and results developed by some RACE projects in the field of service engineering.

ADAM: a problem-oriented symbol processor

Digital computers have evolved in their own technical environment, and to a large degree independently of the problem environment. Thus it was necessary to have computing centers with staffs of programmers as intermediaries between machines and users. As the inadequacy of the arrangement became apparent, problem-oriented languages were written, with compiler programs to allow the machines themselves to do the conversion to their own (machine) language. Accommodating to the nature of the computer in this way still was not the answer from the scientists or experimenters point of view, for there remained an enormous commitment of processing (compiling) and debugging prior to the first feedback of results. Furthermore, it proved necessary to write compilers for many problem fields, which gave this mode of solution a patchwork look. For these reasons, we decided to attack the problem at its roots by changing the fundamental nature, i.e., the organization of the computer itself. We took it as the aim of our work to allow the experimenter to use the computer as directly as possible as an experimental tool.

A processor allocation method for time-sharing

A scheduling algorithm is proposed which is intended to minimize changes of tasks on processors and thereby reduce overhead. The algorithm also has application to more general resource allocation problems. It is implemented by means of a method for efficiently handling dynamically changing segmented lists.

The RACE Open Services Architecture project

The specification and implementation of current telecommunication services tend to be intimately bound to a specific network architecture. Moreover, within the service software, interactions between the logical modules are not always explicit, accessible, or uniform, and tend to be optimized for a particular service. This is exemplified by the difficulty experienced in integrating equipment from multiple vendors, and has resulted in telecommunication systems that cannot rapidly exploit the advantages of new technology or respond to changing customer requirements. In addition, current telecommunication services tend not to be viewed as an integral whole, whereby user, control, and management aspects of a service are developed independently from one another. Separate development can lead to problems of inconsistency if shared data are not updated correctly. The RACE Open Services Architecture (ROSA) project was established to address these problems. This paper presents an overview of the approach taken in the ROSA project.

A Procedure-Oriented Machine Language

A study of the general area of problem solving with a digital computer has revealed characteristics of data that are essentially ignored or suppressed in conventional systems. In an attempt to increase the capability and flexibility of a digital system, a new, high-level language, which utilizes these data characteristics has been defined. A machine organization, which implements this language as a machine language and yet imposes no restrictions on the use of the language has been proposed. This paper presents a general description of the philosophy and structure of the language.

The European telecommunications research and development program RACE and its software project SPECS

This paper presents the RACE program and the objectives and achievements of SPECS, a representative RACE project. The European Commission has set up the research and development program RACE for the preparation and promotion of an integrated broadband communication system in Europe. The SPECS project develops methods and techniques for the development of the complex software needed by this communication system. Its approach is the use of formal methods and maximum automation. A unique feature of this approach is the support of multiple specification languages, including the ability to mix specification languages within a given system design.

Achieving a Pan-European Service Infrastructure

This paper presents a view on what still is to be achieved in the Pan-European Service Infrastructure in order for Europe to have in place an integrated broadband information and communication system. The problems and issues discussed are considered as being of two types: technology and construction. Important technology issues are that the service infrastructure should assure that services are able to be introduced as easily and efficiently as possible, that services are able to cooperate with other services, and that services are not able to interfere with other services. Important construction issues are to provide a support infrastructure, and to assure that a sufficient set of services are created and offered, that the set is well integrated with the different services able to cooperate when desired, that the set is well managed, that the services are usable and able to be accessed, and, finally, that the system can provide value to the users and profit to the service providers. The papers in this book show that most problems and issues dealing with technology are solved. Building the software still presents a formidable challenge, however. It is proposed that a way to ease the challenge is to divide and conquer — to reduce to the minimum the building of universal architectures, but to build services and associated architectures within individual subsets of services, each subset having some community of common interests and characteristics.