Date of Award

Fall 1986

Document Type

Thesis - Restricted

Degree Name

Master of Science (MS)

Department

Mechanical and Industrial Engineering

First Advisor

Chandler, Edward W.

Second Advisor

Schneider, Susan C.

Third Advisor

Josse, Fabien

Abstract

The factory is characteristically a high level ambient electromagnetic interference (EMI) environment. Distributed control communication networks designed to operate in this environment require special design considerations in order to assure that an acceptable level of bit error rate (BER) performance is maintained. Presently, the theoretical BER performance of communication transceivers is typically calculated on the basis of an assumed additive white Gaussian thermal noise model. In order to achieve this theoretically predicted BER performance the EMI susceptibility of the transceivers must be minimized. The problem is then to predict the level of EMI "hardening" necessary to cause the BER performance of the communication transceivers to be primarily determined by the intrinsic thermal noise. This prediction problem is complicated by the fact that real-world ambient EMI is not well-characterized by a Gaussian model. A statistical non-Gaussian model of the interference process developed by Middleton is therefore used to predict when the EMI will be a significant factor in determining the BER performance of the system. Models of the EMI coupling mechanisms and ambient EMI data are then used to estimate the total interference present at the input of a receiver. EMI susceptibility "fixes" can then be applied iteratively to a design until the effects on the BER performance of the coupled EMI is insignificant.

Share

COinS

Restricted Access Item

Having trouble?