Showing papers on "Differentiated service published in 1991"

Distributed configuration profile for computing system

Abstract: A distributed computing system using a data communications network may have a number of service providers for a given service or remote procedure call. A client on the network makes reference to a name service to obtain the network address of one of these service providers. The name service maintains for each client or group of clients a configuration profile of the service providers in order to resolve the issue of selecting one of the several service providers when a request is made. A single configuration profile is a priority-ordered search list that maps from a service identifier (e.g., remote procedure call interface specification) into service provider (e.g., remote procedure call server) names. A configuration profile may include names for individual service providers, and/or named groups of service providers, and/or other configuration profiles. Configuration profiles are stored in a manner that makes them accessible throughout the distributed system, e.g., in the name service. Configuration profiles may be chained together by referencing other configuration profiles to provide a hierarchy of configuration profiles.

Technology, Design and Service Quality

Abstract: Technology as a key dimension of service quality has generally been overlooked. When embedded in organisation culture, technology creates a competitive advantage sustainable over time because it is not easily imitated. The value of developing operational definitions of service quality dimensions is explored. Technology is investigated not only as input‐processing‐output but as the application of knowledge to work. Two components of technology are: service logic (know‐why) and service technique (know‐how). Service logic is offered as the means for expressing causal relationships in a way that establishes rational connections between seemingly unrelated operating events. Service technique is viewed as a way of doing things designed to optimise customer satisfaction while maximising employee satisfaction. The service map is presented as a managerial tool for discovery of the service logic that gives rise to service technique.

Service creation technologies for the intelligent network

Abstract: An intelligent network lets a switch query a network node for help in completing a call. Such a node, for example, can translate an 800 number into a true network address. The network service provider often creates the logic in an intelligent network, but frequently that logic is written by sophisticated subscribers (business customers). Since subscribers can create their own service logic, the service creation system must ensure that such customized logic will not cause the network node to fail or in any way affect the service of other subscribers. Service creation technologies are the tools and technical infrastructure designed specifically to help create and update services. AT&T has created software tools to support application-oriented languages. Its service creation technologies are evolving to meet service provider and service subscriber demands for more control over the services they provide and use.

Customer-defined service model and definition method for intelligent networks

Abstract: The authors propose a customer-defined service model and definition method They also present a visual user interface based on this service model A description is presented of how a customer can easily define a service comprising typical service features like virtual private network (VPN) closed user group (CUG), and private numbering plane (PNP) >

Hierarchical service control scheme for intelligent network services

Abstract: With the introduction of communication services into business activities, such as telemarketing, customer requirements for these services are becoming more varied. Since communication service has become competitive, quick provision of services has become important. This paper proposes a hierarchical service control scheme aimed at achieving service specification independence and the software architecture of an experimental intelligent network service system based on this new scheme. In the proposed new scheme, elementary functional parts are prepared beforehand and a service is realized using interpreter control based on the information about an execution order of functional parts. To localize modifications in software when a new service is introduced, functional parts are arranged in a hierarchical manner. This experimental system can change complicated service specifications in a few minutes. Processing volume is about half of that for an existing system in which service specification is described by a programming language and service associated data are managed by a common data management program.

A network service acquisition mechanism for the client/service model

Abstract: Three of the most important issues in exploiting network servers in existing heterogeneous systems are the problems of service specification, fault tolerance, and system integration. The service specification issue concerns how to specify services so that service-server bindings can be changed dynamically without disturbing clients. The fault tolerance issue concerns how to make clients resilient to network or server failure. The system integration issue concerns how to accommodate server protocol heterogeneity to provide a single system view to the clients. This thesis proposes a service acquisition mechanism to solve these issues. This mechanism is designed under a new computational model called the client/service model. This model extends the client/server model with a new abstract entity to make the notion of service a first-class concept. This new abstract entity appears to the client as a service provider, and to the server as a service requester. This mechanism is evaluated through the design and implementation of a prototype heterogeneous distributed system called Cygnus. While different servers in this system may support different access mechanisms, their interface operations are abstracted as services and specified by an attribute-based naming scheme. They need not be modified to make them accessible to Cygnus clients. We have also designed and implemented a programming language called CygnusC to demonstrate how a typed procedural language like ANSI C can be extended to support the proposed mechanism, though the mechanism can be used within a C program through a set of library routines. The service abstraction facilities provided by CygnusC in the Cygnus distributed system (1) allow programs to specify and access network services based upon their properties, (2) enable the compiler to type check the operations on attribute values, to incorporate users' code for validating service specifications, and to type check every service access operation, and (3) permit the service acquisition runtime to change service-server bindings whenever necessary without disrupting the services in use. Our preliminary results show that it is useful and feasible to embed the Cygnus service acquisition mechanism into a typed procedural language for accessing network services in a large heterogeneous distributed system.

Economic Analysis of Reliability Differentiated Service

Abstract: Welfare geins and reserve reduction of giving Customers the choice of reliability of service is analyzed. The problem of specifying reliability indices to give proper signals to the consumer is outlined. The welfare maximization formulation of Ramsey has been extended to include reliability constraints. The results shou uncertain benefits from reliability-base d pricing. There appears to be no economic compulsions to uarrant such pricing schemes.

Economic analysis of reliability differentiated service (for electricity consumers)

Abstract: Welfare gains and reserve reduction for giving electricity customers the choice of reliability of service are analyzed. The problem of specifying reliability indices to give proper signals to the consumer is outlined. The welfare maximization formulation of Ramsey has been extended to include reliability constraints. The results show uncertain benefits from reliability-based pricing. There appears to be no economic compulsion to warrant such pricing schemes. >