Date of Award

3-14-2014

Document Type

Thesis

Degree Name

Master of Science in Electrical Engineering

Department

Department of Electrical and Computer Engineering

First Advisor

Richard K. Martin, PhD.

Abstract

This research focuses on blind demodulation of a pass band OFDMA signal so that jamming effectiveness can be assessed; referred to in this research as BDA. The research extends, modifies and collates work within literature to perform a new method of blindly demodulating of a passband OFDMA signal, which exhibits properties of the 802.16 Wireless MAN OFDMA standard, and presents a novel method for performing BDA via observation of SC LA. Blind demodulation is achieved by estimating the carrier frequency, sampling rate, pulse shaping filter roll off factor, synchronization parameters and CFO. The blind demodulator's performance in AWGN and a perfect channel is evaluated where it improves using a greater number OFDMA DL symbols and increased CP length. Performance in a channel with a single multi-path interferer is also evaluated where the blind demodulator's performance is degraded. BDA is achieved via observing SC LA modulation behavior of the blindly demodulated signal between successive OFDMA DL sub frames in two scenarios. The first is where modulation signaling can be used to observe change of SC modulation. The second assumes modulation signaling is not available and the SC's modulation must be classified. Classification of SC modulation is performed using sixth-order cumulants where performance increases with the number of OFDMA symbols. The SC modulation classi er is susceptible to the CFO caused by blind demodulation. In a perfect channel it is shown that SC modulation can be classified using a variety of OFDMA DL sub frame lengths in symbols. The SC modulation classifier experienced degraded performance in a multi-path channel and it is recommended that it is extended to perform channel equalization in future work.

AFIT Designator

AFIT-ENG-14-M-65

DTIC Accession Number

ADA600208

Share

COinS