Date of Award


Document Type


Degree Name

Master of Science

First Advisor

Gregg H. Gunsch, PhD


During the past several years, there has been a significant increase in interest in the use of packetized audio over packet-switched networks due to economic and technical feasibility. However, distribution of real-time voice traffic is basically different from traditional reliable data transfer since continuous voice is sensitive to end-to-end delays and variations of delays. The distribution of continuous voice across a packet-switched network requires consideration of encoding schemas, end-to-end network delays, delay variations, and packet loss, all of which significantly affect the playback quality at the receiving side. There is a trade-off among these factors that affects the quality of service. This trade-off is analyzed in this effort. The packet voice system is modeled and analyzed in order to determine this trade-off. Experimental network measurements are accomplished in order to provide realistic inputs to these simulations. The quality of service (QoS) is measured by end-to-end and delay and probability of gap due to late or lost packets in the analysis. Several mathematical expressions of QoS factors and metrics are developed based on simulation results. These mathematical expressions can be used to optimize the system or to estimate the quality of service for a given operating condition.

AFIT Designator


DTIC Accession Number



The author's Vita page is omitted.