COMPUTATIONAL TECHNIQUES FOR THE EVALUATION OF TOTAL SERVICE TIME IN PACKET-SWITCHED STORE-AND-FORWARD COMMUNICATION NETWORKS: CYCLIC ALLOCATION AND BOUNDING PROCEDURES.

Item

Title
COMPUTATIONAL TECHNIQUES FOR THE EVALUATION OF TOTAL SERVICE TIME IN PACKET-SWITCHED STORE-AND-FORWARD COMMUNICATION NETWORKS: CYCLIC ALLOCATION AND BOUNDING PROCEDURES.
Identifier
AAI8222977
identifier
8222977
Creator
RUDOWSKY, IRA STEPHEN.
Contributor
Jacob Rootenberg
Date
1982
Language
English
Publisher
City University of New York.
Subject
Computer Science
Abstract
The research reported herein deals with the problem of computing the total service time of packet-switched messages that are stored-and-forwarded from source node to destination node over computer-communication networks.;Previous results reported on in the literature relied exclusively on the Independent Assumption. The research to be reported on here divorces itself from the Independence Assumption and obtains results regardless of the arrival distribution of messages at the source node. This provides a more realistic solution to the real-world problem.;Two techniques are presented for the computation of total service time--one for an exact solution and one for computing upper and lower bounds. The latter technique, based upon the novel idea of cyclic allocation, complements the former by allowing a quick weeding-out of the infeasible solutions by testing if the required total service time falls within the bounds. This is done in much less time than needed by the exact solution. As the exact solution is a recursive algorithm whose complexity grows exponentially, the bounding techniques also save time and money for the analyst by allowing him to concentrate only on viable configurations. Thus, two original tools are provided to enhance the process of network analysis and design.;Both techniques have been programmed in interactive PL/I and are included in the appendices of the research.
Type
dissertation
Source
PQT Legacy CUNY.xlsx
degree
Ph.D.
Program
Engineering and Computer Science
Item sets
CUNY Legacy ETDs