Date of Award

5-1995

Document Type

Thesis

Degree Name

Master of Science

Department

Department of Electrical and Computer Engineering

First Advisor

Edward F. Mykytka, PhD

Abstract

This study investigates the application of statistical process control methods to monitoring the performance of a communication network. The methods applied include four different types of control charts. The literature search uncovered only one previous study that used a control chart to monitor a communication network. Using a case study of a communication network, four important issues for proper control chart usage are emphasized. These issues are: proper data collection rate due to autocorrelation, proper subgrouping of the data, ensuring that count data conforms to the assumptions of the binomial probability model before implementing p or np control charts, and viability of using subgroups of aftribute data as measurement data on an x-bar chart. The results indicate that control charts are indeed a viable method for monitoring a communication network's performance over time, especially when the available data on the network is limited.

AFIT Designator

AFIT-GOR-ENGS-95J-01

DTIC Accession Number

ADA305757

Download

Share

COinS