Proceedings

Consider a queueing system in which customers (or jobs) arrive to one of Q separate queues to await service from one of S identical servers (Figure 1). Once a job enters a queue it does not leave that queue until it has been selected for service. Any ...

The problem of optimal control of a M/M/m queueing system is investigated. As in the M/M/l case the optimum control is shown to be a window type mechanism. The window size L depends on the maximum allowable time delay T and can be explicitly computed. ...

In this paper, we propose a class of queueing network models, and a method for their approximate solution, for computer networks with bursty traffic and finite buffer space. The model is open, implying no population limit except for buffer size limits ...

Packet switching networks with flow-controlled virtual channels are naturally modeled as queueing networks with closed chains. Available network design and analysis techniques, however, are mostly based upon an open-chain queueing network model. In this ...

Three different load balancing algorithms for distributed systems that consist of a number of identical processors and a CSMA communication system are presented in this paper. Some of the properties of a multi-resource system and the balancing process ...

In the design and implementation of computer networks one must be concerned with their overall performance and the efficiency of the communication mechanisms chosen. Performance is a major issue in the architecture, implementation, and installation of a ...

Application and system software architecture can greatly influence the operational statistics of a local network. The implementation of a transparent file system on top of a high bandwidth local network has resulted in generating a high degree of file ...

SNA is both an architecture and a set of products built in conformance with the architecture (1,2,3). The architecture is layered and precisely defined; it is both evolutionary and cost effective for implementing products. Perhaps the largest component ...

As part of its operation and maintenance of the ARPANET for the past twelve years, BBN has been asked to investigate a number of cases of degradation in network performance. This presentation discusses the practical methods and tools used to uncover and ...

This paper deals with the economic tradeoffs associated with data compression in a packet switching environment.

In section II we present the data profile concept and the compression analysis of typical file-transfer data strings. This is followed by a ...

A general model of Intelligent Communications Concentrating Devices (ICCD) is presented and analyzed for delay and overflow performance with compensation for the pipelining effect of message arrivals extending over time. The results of the analysis ...

We consider a special class of homogeneous computer network comprising several essentially identical but independent computing systems (ICSs) sharing a single resource. Of interest here are the effects of modularly expanding the network by adding ICSs. ...

Most of the analysis of buffer sharing schemes has been aimed at obtaining the optimal operational parameters under stationary load situations. It is well known that in most operating environments the traffic load changes. In this paper, we address the ...

We analyze the performance of a window multiaccess protocol with collision size feedback. We obtain bounds on the throughput and the expected packet delay, and assess the sensitivity of the performance to collision recognition time and packet ...

The communication loops analyzed here provide an economic way of attaching many different terminals which may be some kilometers away from a host processor. Main potential bottlenecks were found to be the loop transmission speed, the loop adapter ...

Diverse network application requirements bring about local networks of various size, degree of complexity and architecture. The purpose of this paper is to present a network protocol layer which is used to provide a homogeneous operating environment and ...

Performance measurement and analysis of the behavior of a computer network usually requires the application of multiple software and hardware tools. The location, functionality, data requirements, and other properties of the tools often reflect the ...

This paper describes the Hierarchical Modeling System (HMS). HMS is a tool - a unified and expandable system - which supports the development of analytic and simulator models of computer networks. The same system and workload descriptions can be ...

Managing networks using Network Administration Centers is increasingly considered. After introducing the information demand for operational, tactical and strategic network management the paper is dealing with the investigation of the applicability of ...

Much has been written and spoken about the capabilities of emerging designs for Local Area Networks (LAN's). The objective for this panel session was to gather together companies and agencies that have brought LAN's into operation. Questions about the ...

The communication loops analyzed here provide an economical way of attaching many different terminals to a IBM 4331 host processor which may be several kilometers away.

As a first step of the investigation protocol overhead is derived. It consists of ...

At Bethesda, Maryland, the National Library of Medicine has a large array of heterogenous data processing equipment dispersed over ten floors in the Lister Hill Center and four floors in the Library Building.

The National Library of Medicine decided to ...