Showing papers in "Ibm Systems Journal in 1992"

Extending and formalizing the framework for information systems architecture

Abstract: John Zachman introduced a framework for information systems architecture (ISA) that has been widely adopted by systems analysts and database designers. It provides a taxonomy for relating the concepts that describe the real work to the concepts that describe an information system and its implementation. The ISA framework has a simple elegance that makes it easy to remember, yet it draws attention to fundamental distinctions that are often overlooked in systems design. This paper presents the framework and its recent extensions and shows how it can be formalized in the notation of conceptual graphs.

A global approach to crew-pairing optimization

Abstract: The problem addressed in this paper is crew-pairing optimization in airline flight planning: finding tours of duty (pairings) that are legal and cover every flight leg at the least cost. The legal rules and cost of a pairing are determined by complex Federal Aviation Agency and contractual requirements. In addition, the problem is made more difficult by the hub-and-spoke system used by airlines that multiplies the possible ways a pairing can link flight legs. The state-of-the-art crew-pairing TRIP system of American Airlines uses subproblem optimization and, as is true for other crew-scheduling systems, may not be able to improve a solution even though a better one exists. We report on the methodology developed during a joint study by IBM and American Airlines Decision Technologies to use the IBM Optimization Subroutine Library in conjunction with TRIP to improve on crew-pairing solutions by taking a global approach. The resulting improvements have been a reduction of 5 to 11 percent in excess crew cost. Estimated total savings are five million dollars per year.

Interactive image segmentation for radiation treatment planning

Abstract: COVIRA (COmputer VIsion in RAdiology) is a project in the European Community's Advanced Informatics in Medicine program. The goal is to improve the diagnosis and planning of treatment (radiotherapy) for patients with brain tumors and other diseases. The aim of radiotherapy is to provide a high dose of radiation to a tumor while sparing as much as possible of the surrounding healthy tissue. A necessary first step is defining the target volume and organs at risk by manually outlining the required contours on magnetic resonance or computed tomography scans. For a full three-dimensional plan this is time-consuming, as 40 or more scans are used. Computer image segmentation speeds up the process, and a method that combines information from edge and region detectors is described. Since this method is not able to completely meet the clinical requirements, an interactive image segmentation algorithm has been developed that enables the operator to employ clinical judgment. Probabilities are assigned to edges and regions and presented to the user as a hierarchy of segmentations. The approach is being subjected to extensive clinical evaluation, using pilot applications running on IBM RISC System/6000 workstations.

Casual probabilistic network modeling: an illustration of its role in the management of chronic diseases

Abstract: This paper describes the role of the novel technique of causal probabilistic network (CPN) modeling as an approach to tackling control system problems typified by that of the administration of treatment to the patient suffering from a chronic disease such as diabetes. Three roles of a CPN are discussed. First, since diabetes arises as a consequence of impaired control of carbohydrate metabolism, the ability of a CPN to represent the uncertainty of a physiologically-based model is described. Second, its ability to make robust estimates of the parameters of the metabolic model is presented, and finally, in conjunction with decision theory approaches, its ability to compare alternative therapies and advise on insulin therapy for patients with insulin-dependent diabetes mellitus is illustrated.

Frontier: a graphical interface for portfolio optimization in a piecewise linear-quadratic risk framework

Abstract: “Frontier” is a pilot graphical user interface for portfolio optimization built for the new IBM workstation, the RISC System/6000™, out of basic X-windows and OSL utilities. The program asks the user to select a piecewise linear-quadratic risk measure, draws a risk/reward efficient frontier, and permits the user to examine the efficient frontier using zoom and histogram display facilities. This paper describes the interfaces and discusses possible extensions.

Recent developments and future directions in mathematical programming

Abstract: Recent advances in mathematical programming methodology have included: development of interior methods competing with the simplex method, improved simplex codes, vastly improved performance for mixed-integer programming using strong linear programming formulations, and a renewed interest in decomposition. In addition, use of vector and parallel processing has improved performance and influenced algorithmic developments. Application areas have been expanding from the traditional refinery planning and distribution models to include finance, scheduling, manufacturing, manpower planning, and many others. We see the acceleration of better methods and improved codes moving together with faster, lower-cost, and more interesting hardware into a variety of application areas, thereby opening up new demands for greater function of optimization codes. These new functions might include, for example, more powerful nonlinear codes, decomposition techniques taking advantage of network and other problem-dependent structures, and mixed-integer capability in quadratic and general nonlinear problems.Stochastic scenario programming and multitime-period problems are becoming solvable and open up applications and algorithmic challenges. The IBM Optimization Subroutine Library has helped to accelerate these changes but will have to continue to change and expand in ways that are touched upon in this paper.

Implementing interior point linear programming methods in the Optimization Subroutine Library

Abstract: This paper discusses the implementation of interior point (barrier) methods for linear programming within the framework of the IBM Optimization Subroutine Library. This class of methods uses quite different computational kernels than the traditional simplex method. In particular, the matrices we must deal with are symmetric and, although still sparse, are considerably denser than those assumed in simplex implementations. Severe rank deficiency must also be accommodated, making it difficult to use off-the-shelf library routines. These features have particular implications for the exploitation of the newer IBM machine architectural features. In particular, interior methods can benefit greatly from use of vector architectures on the IBM 3090™ series computers and “super-scalar“ processing on the RISC System/6000™ series.

Introduction to the IBM optimization subroutine library

Abstract: This essay introduces the IBM Optimization Subroutine Library (OSL) and seven OSL-related papers that appear in this issue. Developed as a result of a partnership between several IBM research and development groups, OSL provides a suite of tools for manipulating the models and solving the resulting minimization and maximization problems of mathematical optimization. The problems that OSL addresses include: linear, quadratic, mixed-integer, and pure network programming problems. OSL includes solvers based on the classical simplex method and on newer interior point methods. Because a user-supplied driver program coordinates the problem solution, and because of the “mix and match” philosophy of OSL, a user may, within rather wide limits, individually tailor a technique to solve a particular problem. We conclude that OSL is something new in optimization software.

Project Athena: supporting distributed computing at MIT

Abstract: Project Athena™ was an educational computing initiative at the Massachusetts Institute of Technology, undertaken in partnership with the IBM Corporation and Digital Equipment Corporation from 1983 to 1991. This paper gives an overview of the network-based distributed computing services, developed for a number of UNIX™-capable workstations. These services are extensions to the native operating systems of the workstations, and provide interoperability as well as systems administration facilities in a large heterogeneous workstation environment. Under Project Athena, a mature distributed computing environment was developed. Its organization and support structure may be used as a model when planning a new installation, whether on a university or commercial campus. A section of this paper deals with the support requirements for distributed computing environments, based on the Project Athena experience.

The Open Document Architecture: from standardization to the market

Abstract: The Open Document Architecture (ODA) was developed in the mid-1980s by several standardization bodies. It is now a stable set of international standards for the interchange of compound documents consisting of text, image, and graphic content. Since 1985 the standardization process has been accompanied by European industrial cooperation projects in order to get early experience with the standard and to develop technologies implementing the standard. IBM's European Networking Center has participated in the projects and has prototyped enhancements to OfficeVision™ platforms to allow the interchange of ODA documents between OfficeVision applications and applications running on other vendor platforms. In this paper, the ODA technology is described, experiences of interworking in heterogeneous environments are given, and the role of cooperating with project partners is outlined.

A decomposition method for quadratic programming

Abstract: We discuss the algorithms used in the Optimization Subroutine Library for the solution of convex quadratic programming problems. The basic simplex algorithm for convex quadratic programming is described. We then show how the simplex method for linear programming can be used in a decomposition crash procedure to obtain a good initial basic solution for the quadratic programming algorithm. We show how this solution may be used as a starting solution for the simplex-based algorithm. Besides its ability to obtain good starting solutions, this procedure has several additional properties. It can be used directly to find an optimal solution to a quadratic program instead of simply finding a good initial solution; it provides both upper and lower bounds on the objective function value as the algorithm proceeds; it reduces the complexity of intermediate calculations; it avoids certain numerical difficulties that arise in quadratic, but not linear programming

Naming and registration for IBM distributed systems

Abstract: Today's trends toward interconnection of networks expose limitations and deficiencies of traditional identification schemes. The need arises for a uniform naming solution that can accommodate the size and heterogeneity of worldwide domains, while still remaining understandable, usable, and manageable by human users. This paper describes a proposatol name objects and resources in distributed environments; each object or resource cabne located, accessed, communicated with, operated on, managed, or secured using the same, unique name. The solution proposed here includes registration mechanisms necessary to ensure name uniqueness. The schemeis based on existing standards, mainly Open Systems Interconnection (OSI) Distinguished Names; whenever standards disagree, the preference goesto the alternative that offers the widest usage acrosasll protocols. Clear, consistent naming guidelineasr e given that would enable ISM customers who have purchased IBM networking and system management products to name their resourcesso that their administrative processes and IBM's products and protocols can support those resources effectively. A method is suggested to encompass existing name spaces in a single, worldwide namings pace, and a migration pathis sketched. The interoperation of different protocols across network boundariesu sing the same naming constructs is shown by means of several scenarios. The naming and addressing scheme proposed here requires nothing new or different from the already defined standards, but allows interoperation among them by using a subset of each.

APPC/MVS distributed application support

Abstract: Advanced Program-to-Program Communication for Multiple Virtual Storage (APPC/MVS) is a major evolutionary change to MVS for applications that need to connect and communicate across the enterprise. APPC/MVS is an implementation of Systems Network Architecture (SNA) LU 6.2 session-defined protocol. This new MVS environment includes services to enhance the creation, execution, and management of MVS peer-to-peer and client/server applications. In addition to providing connectivity and communications services, APPC/MVS also has scheduling facilities for managing concurrent work originating from other systems in the enterprise network. The objective of this paper is to survey these facilities by examples of usage and by relationships to existing MVS services. Topics include various types of distributed models and approaches, design considerations, and characteristics that represent candidates for an APPC/MVS implementation. Relationships of the APPC/MVS environment with the Customer Information Control System/Enterprise Systems Architecture (CICS/ ESA™), Information Management System/ Enterprise Systems Architecture Transaction Manager (IMS/ESA™ TM), Time Sharing Option Extensions (TSO/E), and batch environments are included.

Service and traffic management for IBCN

Abstract: The future Integrated Broadband Communications Network (IBCN) will provide high-speed communication capabilities that support a variety of existing and new services. The management of such a complex environment requires innovative management systems. NEMESYS is a project within the European Commission's Research and Development in Advanced Communications in Europe (RACE) program. The project goals are to demonstrate and evaluate the use of advanced information processing techniques for quality-of-service and traffic management. To reach these goals, a series of experimental prototypes are being built. This paper describes the assumptions, objectives, and approach of NEMESYS, and in particular, the design and implementation of an experiment that investigates service and traffic management techniques in a simulated asynchronous transfer mode environment. Because the project is not yet finished, some preliminary results are presented.

Architectural directions for opening IBM networks: the case of OSI

Abstract: This paper discusses the results of a research project that developed an architectural framework for integrating non-IBM network architectures to the reference model and node structures of IBM's Systems Network Architecture (SNA). The unique features of the selected integration approach allow multiple protocol stacks to coexist and interoperate within the same computer, to share use of common physical network ports, links, and switching nodes, and to be accessed and managed through homogeneous interfaces. The architectural framework was developed for the specific purpose of integrating the Open Systems Interconnection (OSI) Reference Model to that of SNA, but its basic philosophy and key aspects turn out to be generally applicable to the integration of other network technologies as well, such as TCP/IP or NetBIOS.

RODM: a control information base

Abstract: Operational management of computers and computer networks was formerly performed exclusively by an operator or a team of operators equipped only with consoles for the display of status messages. Each system component independently determined its own set of such messages, identifying conditions needing attention. To meet future challenges, however, a structured approach to systems and network management and associated automation will be necessary. The amount and complexity of the status information needed for control and coordination will make it unlikely that operators will be able to keep up with such needs unaided. This control information must be made available to a family of systems and network management applications (including operator display programs). The NetView® Resource Object Data Manager (RODM) is designed to facilitate the storage and retrieval of control information. It provides services for defining a structured data model of a computer system. The control information is not kept simply in the form of messages, but instead the data are organized into units called objects. This allows the model to effectively capture interrelationships and dependencies as well as status information.

The evolution of the Common User Access workplace model

Abstract: This paper describes some of the influences contributing to and issues in dealing with the evolution of user interface guidelines over time. In particular, we focus on the evolution of IBM's user interface architecture, the Common User Access™ (CUA™) interface, over a period of six years. Discussed are the key architectural and design elements of the CUA Workplace Model, the fundamental shifts in computer-human interaction that have occurred since the first publication of the guidelines in 1987, and how user interface design, operating systems, and tools have interacted in the evolution of the guidelines. The information should help designers of user interfaces and developers of useirn terface guidelines to appreciate some of the factors involved in the long-term evolution of a user interface style. The paper provides an introduction to the most recent evolutionary step in the CUA style (the Workplace Model) to help the reader place these factors in perspective relative to the degree of evolutionary change.

The RACE Open Services Architecture project

Abstract: 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

Design considerations for distributed applications

Abstract: Probably the hardest part about developing a distributed application is determining where to start. There are multiple hardware and software platforms to understand, network traffic implications, and numerous tools and technologies to consider. One question, however, transcends the importance of what platform to pick or what tool to use: that is, how do you design it? This paper represents the results of two years of work with customers regarding this question. The paper explores some of the implications of working in a distributed environment, reviews some rules for data and function placement, and introduces a methodology for distributed application design.

AIX NetView/6000

Abstract: AIX® NetView®/6000 is a network management system that manages simple network management protocol (SNMP) devices developed by IBM and other vendors. It provides configuration, fault, and perfor-mance applications integrated into an advanced end-user interface (EUI), which incorporates a graphic display of network topology and performance as well as system management functions accessible from both graphic and character-based devices. An application builder and event configurator allow users to generate performance applications and provide automation of management tasks specific to their networks. In addition to providing stand-alone distributed network management, AIX NetView/6000 also provides a bidirectional connection to IBM's mainframe-based NetView product to enable central management of the enterprise network from System/370™ and System/390™ NetView.

Evolution of an open communications architecture

Abstract: An overview of the current IBM communications paradigm for interconnecting computer networks is presented. Emphasis is on the incorporation of multiprotocol, multivendor facilities in an integrated architecture. This paper presents an overview of key elements of the evolving communications architecture.

Role of the DASD storage control in an Enterprise Systems Connection environment

Abstract: This paper compares the Enterprise Systems Connection Architecture\" (ESCOflf\"), with its use of fiber optic cables, to the parallel channel architecture introduced with System/360TM. It also describes many of the reasons for the introduction of ESCON. The ESCON implementation for the IBM 3990 Storage Control is described in some detail, including a description of nonsynchronous operation. The paper concludes with a discussion of some of the benefits of ESCON for a 3990 installation, performance considerations, and migration considerations.

SNA management services architecture for APPN networks

Abstract: The introduction of Advanced Peer-to-Peer Networking (APPN) provides for a more flexible Systems Network Architecture (SNA) environment: end-user systems (physical units, or PUs) at the edge of a routing network no longer need a predefined relationship with a system services control point (SSCP) for network control purposes. This new flexibility creates challenges for SNA Management Services, however, since the SSCP-PU relationship provided a vehicle for network management as well as network control. To meet the needs of this peer-to-peer environment, the SNA Management Services architecture was extended to provide a management infrastructure that replaces the previous SSCP-PU relationship, and at the same time provides for much greater flexibility. This new infrastructure consists of a formalization of the focal-point/entry-point concept in the architecture and a transport technique for management services data that utilizes the facilities of Advanced Program-to-Program Communications (APPC) rather than the SSCP-PU session. Together this provides for a management structure in a peer network.

Data description and conversion architecture

Abstract: A data description and conversion architecture has been defined by IBM to enhance data interchange among Systems Application Architecture™ (SAA™) programming languages and systems. Its components, described in this paper, are (1) A Data Language (ADL), a programming language for describing data and specifying what data conversions are to be performed, (2) an object-oriented method of encoding ADL for efficient machine storage, transmission, and processing, and (3) programs that translate the data declarations of other programming languages to or from ADL. Also discussed is the application of the architecture to record-oriented files for SAA Distributed File Management.

A common compiler for LOTOS and SDL specifications

Abstract: This paper describes a translation of LOTOS and SDL specification languages into executable code, as it was prototyped in the Specification and Programming Environment for Communication Software (SPECS) project under the Research and Development in Advanced Communications in Europe (RACE) program. Both languages are translated into a common intermediate representation in the form of a network of state machines with both synchronous and asynchronous communications. By a series of transformations that make full use of the equivalence relations defined on LOTOS processes, this translation solves unique problems stemming from the highly abstract nature of LOTOS. The common intermediate representation is mapped into C code that can be executed in a specific run-time environment, implemented on a UNIX®-like operating system. SPECS has also developed a pragmatic approach to represent implementable data types in the algebraic framework of LOTOS and SDL, based on a set of predefined type constructors.

NetView Version 2 Release 3 Graphic Monitor Facility: network management graphics support for the 1990s

Abstract: The NetView® Version 2 Release 3 Graphic Monitor Facility provides an Operating System/2® (OS/2®) workstation-based graphics user interface for Net-View that permits an operator to view graphically and control via generic commands both Systems Network Architecture (SNA) and non-SNA networks. This paper describes the major new capabilities of the NetView V2R3 Graphic Monitor Facility, discusses its structure at a high level, and describes in some detail the new support in this release for graphic views of various types of networks, simplified commands for controlling these networks, and the manner in which non-SNA networks are supported.

Network and system automation and remote system operation

Abstract: The rapid growth in the size and complexity of today's information system networks highlights the importance of automation and remote system operation in managing these networks. Version 2 Release 2 of the NetView® program provides improved facilities in these two areas to assist enterprises in consolidating their operations staff in central locations and managing their information systems more efficiently and reliably. This paper describes the Systems Network Architecture (SNA) framework for remote system operation and the NetView remote operations platform, including the LU 6.2 data transport mechanisms and related management services applications. Extensions to the NetView automation platform that permit automation based directly on the receipt of SNA alerts or other structured data are described. In addition, enhancements that improve the performance, usability, and functional capabilities of the NetView automation table are discussed.

Estimating the fault rate function

Abstract: Paging activity can be a major factor in determining whether a software workload will run on a given computer system. A program's paging behavior is difficult to predict because it depends not only on the workload processed by the program, but also on the level of storage contention of the processor. A program's fault rate function relates storage allocation to the page fault rate experienced while processing a given workload. Thus, with the workload defined, the fault rate function can be used to see how the program's storage allocation is affected by varying levels of storage contention, represented by varying fault rates. This paper presents a technique to represent program workloads and estimate the fault rate function, and describes how these results can be used in analyzing program performance.