Search

This research was directed to initiate a study of large-scale hereditary/non-hereditary control and distributed systems under randomly varying structural perturbations. There were three principal areas of research: (1) stability, convergence, etc., (2) error estimation between the stochastic and corresponding deterministic systems, and (3) investigation of applications from various disciplines.

Backup and recovery methods using magnetic tapes are common in computer utilities, since information stored on-line is subject to damage. The serial access nature of the tape medium severely restricts the flexibility and simplicity of accessing and managing the stored data. A method using a data network is described, to present a backup mechanism which takes advantage of a large, inexpensive, random access remote data storage facility to provide data access and management functions that are more flexible than those provided by a traditional backup facility. Although data transfer rates will be reduced, data access and management will be simplified, and system availability will be improved. The work described is based on a network backup facility built for the Multics computer utility, using the ARPAnet.

This paper presents a survey of current research in the area of distributed systems and discusses some possible areas of future research. The trend toward data base management systems which are responsible for the creation, accessing, and maintenance of a large collection of information containing complex interrelationships is prevalent in centralized systems today. It seems that the same advantages which led to the growth of data base management systems in centralized systems can be applied in distributed systems as well. Most current research (particularly in the area of modeling) views a distributed system in terms of independent data files. The authors address some of the problems involved when the distributed system is viewed in terms of an integrated data base.

The purpose of Distributed Processor/Memory Architecture (DP/M) design program was to extend the DP/M avionic system processing concept to a detailed system hardware and software design. The functional design for the DP/ M Processing Element (PE) is summarized, including the processor, memory, input/ output interface, and a dual-level time-division-multiplex bus interface unit. A set of simulation and analysis programs was developed for modeling both the high-level network interaction among interconnected Processing Elements and the detailed internal operation of the PE. Other major areas examined were the executive control software, the process construction methodology required to develop and allocate real-time software for DP/M, and methods that could be used with DP/M to promote avionic system fault tolerance.

The Stanford Network Graphics Project has the goal of providing high-quality interactive graphics over both local-area and long-haul networks. Specifically, a user sitting at an intelligent workstation should have simultaneous access to a variety of graphical and non-graphicl applications distributed throughout an internetwork. Interaction with these applications must be responsive, which, requires that much of the interaction be handled by the workstation itself. To do so the workstation must deal in terms of high-level objects, rather than graphical output primitives. That is, it must provide both modeling and viewing facilities in contrast to contemporary graphics systems. This paper describes the system architecture we have developed and the hardware and software components we are using to realize this architecture in the Standford University Network environment. (Author)

This report summarizes the research performed by the MIT Laboratory for Computer Science from July 1, 1978 to June 30, 1979. (Author)

Distributed sensor systems are key ingredients in many real world applications. Specific instances abound both in industrial and military environments, e.g., the monitoring (and possibly control) of manufacturing operations, or the dispersion of various types of sensors to detect enemy movements. There are two major areas of study in distributed sensors: the design and development of the sensors themselves, and the logical use of such sensors. This report is directed toward mechanisms to study the latter. Some examples of the latter, explained in detail, are based on CICE/ECE/IOE 469, a course in Real Time Computing Systems developed at the University of Michigan, Ann Arbor. Both student and faculty critiques of the electromechanical analog facility used in the program are included., Prepared in cooperation with VerChron Systems, Saline, MI.

The Distributed Master Control Program (DMCP) and the technical control application programs of the Modular System Control Development Model (MSCDM) are described. Program listings of each program and subprogram are presented. Programs are written in FORTRAN and MACRO-11 for Digital Equipment Corporation (DEC) LSI-11 microcomputers., See also Book 2, AD-A078 392.

Program descriptions and listings of the Modular System Control Development Model utility programs and diagnostics are presented. This includes the Bootstrap Load Facility for the Ring Architecture., See also Book 3, AD-A078 393.

Program flow charts for each Modular System Control Development Model (MSCDM) program is presented. Flow charts for all operating systems, application, diagnostic and utility programs are included in this single volume., See also Book 1, AD-A078 391.

This report describes an ongoing project in the area of design of distributed systems. The goal is to develop an effective programming system that will support well-structured design, implementation, maintenance and control of distributed processing applications. This programming system combines a powerful high level language and operating system features, and addresses the underlying system problems that affect the reliability and security perceived on the application level. The report presents a conceptual model of distributed computation, and, in the context of this model, discusses our approaches to inter-node communication and cooperation, reliability, and protection. One of the basic goals of our project is to allow the application programmer to work with application-oriented entities. Thus, inter-node messages, error handling and protection constraints will all be expressible in application oriented terms. The report concludes with some examples of the language constructs and an outline of the future research under this project. (Author)

Under a grant from the Army Research Office, Kansas State University investigated and studied the feasibility of establishing the multiple minicomputers of the School of Aerospace Medicine into a network of computer resources. A PDP 11/70 and a PDP 11/34 were acquired for use as host computer in a distributed processing configuration. Simulations were written to test the network concept prior to installation. The report describes the result of the investigation. (Author)

In this dissertation, a new approach to the synchronization of accesses to shared data objects is developed. Traditional approaches to the synchronization problems of shared data accessed by concurrently running computations have relied on mutual exclusion -- the ability of one computation to stop the execution of other computations that might access or change shared data accessed by that computation. Our approach is quite different. We regard an object that is modifiable as a sequence of immutable versions, each version is the state of the object after an update is made to the object. Synchronization can then be treated as a mechanism for naming versions to be read and for defining where in the sequence of versions the version resulting from some update should be placed. In systems based on mutual exclusion, the timing of accesses selects the versions accessed. In the system developed here, called NAMOS, versions have two component names consisting of the name of an object and a pseudo-time, the name of the system state to which the version belongs. By giving programs control over the pseudo-time in which an access is made, synchronization of accesses to multiple objects is simplified.

The 'architecture' of a computer system defines the interface that the hardware presents to the software of the system, and determines how this interface is realized by subunits of the computer system. In the early days of the stored program computer, knowledge of logic design, the technology of logic and memory devices, elementary machine language programming techniques, and a good measure of common sense were all that was required to be a computer architect. Now things have changed. Our concept of what we expect a computer system to do has reached a high degree of sophistication -- large data bases, multiple concurrent processes, and programming languages that offer recursive programming, abstract data types, protection, and access control. These expectations have been met by elaborate software systems -- operating systems, data management systems and runtime support for language implementations. The ability of system designers to meet these expectations, and the quality of the facilities they can provide to the application programmer, are critically dependent on properties of mechanisms built into the hardware. Thus it has become essential that contemporary computer architects be aware of how architectural decisions interact with software quality -- how hardware structures can more effectively meet the needs of operating systems and modern concepts or program and data structure.

To present the Navy's interest in distributed computer systems, this paper first discusses the nature of Navy systems and some of the constraints created by Navy philosophy. Secondly, some of the requirements of naval systems are presented. A couple of busing systems being investigated by the Navy are mentioned briefly, and some of the problems to be faced, as the author views them, are presented. (Author), Presented at the Computer Software and Applications Conference (COMPSAC 78), Chicago, IL. 13-16 Nov 78.

Results of a workshop held for the purpose of discussing the impact, payoff, and technical issues of distributed computer systems are reported. At the workshop, attendees with diverse backgrounds and experience were able to (1) have a common understanding of distributed systems designs criteria; (2) conduct a quick and complete review of distributed concepts and approaches used in current Navy systems implementation, (3) conduct a state-of-the-art survey of distributed systems theory, (4) compare common issues in relation to Navy-distributed systems, and (5) provide discussions beneficial to both technical and management areas on trends, ideas, and technologies in Navy systems development. (Author)

This report describes BBN efforts in the design of the National Software Works system and BBN efforts to integrate TENEX into the National Software Works system. BBN participation in the National Software Works project is divided into three major areas. These are system design, system implementation, and contractor/sponsor meetings to coordinate design and implementation goals and strategies. During this quarter most of the effort has been in the latter two areas., See also report dated Jun 1976, ADA025956.

This report describes a design concept for an automated distributed data acquisition system. Particular emphasis is placed on uses of automated state-of-the-art data acquisition techniques for monitoring of communication systems. Use of distributed processing, industry standard interfaces, and a real-time software operating system is discussed.

Considerable research is currently going on into the application of distributed computing systems. They appear particularly suitable for the computing needs of a small warship. The particular constraints of the warship's environment are discussed. This is followed by a description of how a ring structured distributed computing system might be adapted to function in this environment. Included in this consideration are the feasibility of attaining adequate bus speed, the use of multiply addressed messages, and methods of handling real-time processing. Of particular interest is the ability to achieve controlled degradation of performance under failure, especially failure due to battle damage.

This report describes the Basic Operating System (BOS) for the Distributed Computer Network (DCN). The BOS is a multiprogramming executive providing process and storage management, interprocess communications, input/output device control and application-program support. It operates with any PDP11 model including at least 4K of storage, an operator's console and a communication device for connection to the DCN. Included in this report is a description of the various components that make up the BOS and the manner in which they operate. Also described are the various primitive functions and command operations used to control the operation of the network and the various application programs. Other reports, listed in the references, describe the functioning of the DCN as a whole and also the upwardly-compatible Virtual operating System (VOS) developed for PDP11 models with memory management features.

The goal of this research in distributed computing is to evolve fundamental concepts as well as implementation techniques for improving system performance by judicious runtime distribution of functions to processors and to levels of a hierarchical operating system. This distribution of computing tasks among the various processors in a network or among the many levels of a hierarchical operating system in a single computer is a critical issue in current and future military, industrial, and educational computer systems. The present approach is to provide the applications programmer with tools that allow both the estimation of performance improvements of potential migrations, either horizontal (inter-processor) or vertical (intra-processor), and the implementation of the most promising migrations., Continuation of Contract No. N00014-67-A-0191-0023.

This report summarizes studies of local, regional, and large scale data communication network problems. Primary emphasis is placed on system issues and tradeoffs. Areas discussed are packet radio system studies, packet radio system algorithms and control, local and regional data network cost comparisons and alternatives, and integrated large scale packet switched network cost and performance. Studies of multidropped, point-to-point, and broad band cable and radio broadcast systems as well as the impact of satellites on network cost and performance are described., See also report dated Jun 1974, ADA016577.

The problems and opportunities involved with designing and using distributed systems are discussed. This is followed by presenting a paradigm for the distributed system design process. The paradigm is then applied to the design of a distributed operating system for SPLICE. The major interface between user and operating system is provided by the dictionary/directory system. (Author)

The ultimate goal of the DAI project is to discover the essential principles upon which construction of a network of cooperative intelligent computer agents could be based. The motivations for distributed AI arise from at least three sources. First, many military problems have an inherently distributed character by virtue of the physical distribution not only of military units themselves but also of the various sensors that provide information affecting those units. Second, the changing economics of computer hardware and communications technology are making distributed systems increasingly cost-effective; since distributed systems with computationally powerful processors will soon be more commonplace, it is worthwhile considering how AI could be used to exploit such systems more fully. Third, there is a strong motivation from within AI. Artificial-intelligence systems that act in an environment must reason about that environment. Thus, when the environment contains other agents, as most realistic environments do, the AI system must reason about the state of those agents and how to affect it. This leads naturally to distributed AI.

The problems and opportunities involved with designing and using distributed systems are discussed. This is followed by presenting a paradigm for the distributed system design process. The paradigm is then applied to the design of a distributed operating system for SPLICE. The major interface between user and operating system is provided by the dictionary/directory system., supported by the Fleet Material Support Office, http://archive.org/details/distributedopera00schn, N0036783POM3883, NA

Aked S., Bolan C., Brand M. (2014). Determining what characteristics constitute a darknet. Proceedings of the 11th Australian Information Security Management Conference, ISM 2013. (pp. 12-20). Edith Cowan University. Available here

This is an Author's Accepted Manuscript of: Baig, Z. A., Sait, S. M., & Binbeshr, F. (2016). Controlled Access to Cloud Resources for Mitigating Economic Denial of Sustainability (EDoS) Attacks. Computer Networks, 97, 31-47. Original article vailable here.

The models of peer to peer protocols are presented with the help of one-step processes. On the basis of this presentation and the method of randomization of one-step processes, it is described method for constructing models of peer to peer protocols. The models of FastTrack and Bittorrent protocols are studied by means of proposed method. © 2014 IEEE.