Statistical Methods of Bit Error Rate Estimation in Optical Communication Systems

This thesis describes two methods of bit error rate (BER) estimation based on statistical techniques. The first method examines the correlation between the BER and electrical characteristics of an optical receiver circuit. Low correlation is found between the receiver performance and the electrical characteristics. This may be caused by a high variability in the quality of the optical portion of the receiver as compared to the electrical portion. In the second method of BER estimation, a preselected analytic function was fit to samples of a signal and this function was used to calculate the probability of error of this signal. The analytic function modeled noise contributions from thermal and modal noise sources. Modal noise, caused by the interference of fiber modes at a surface, is modeled by a gamma distribution when a large number of modes are excited in the fiber. However, if a small number of modes are excited, the gamma distribution does not describe the observed noise distribution. A numerical model is presented which describes the noise distribution in this restricted mode-launch case. This represents the first time noise measurements of restricted mode-launch fiber optic links have been presented.

Name thesis:degree_name

Ph.D.

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Electrical Engineering

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2015

Author dc:creator

• Murphy, George Matthew

Contributors dc:contributor

• George C. Papen

• Statistics

Language dc:language

eng

Identifier

(MiAaPQ)AAI9944944

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/81296