Date of Award

3-26-2020

Document Type

Thesis

Degree Name

Master of Science in Electrical Engineering

Department

Department of Electrical and Computer Engineering

First Advisor

Peter Collins, PhD

Abstract

The development of low-complexity, lightweight and low-cost Non-Destructive Evaluation (NDE) equipment for microwave device testing is desirable from a maintenance efficiency and operational availability perspective. Current NDE equipment tends to be custom-designed, cumbersome and expensive. Software Defined Radio (SDR) technology, and a bandwidth expansion technique that exploits a priori transmit signal knowledge and auto-correlation provides a solution. This research investigated the reconstruction of simultaneous SDR receiver instantaneous bandwidth (sub-band) collections using single, dual and multiple SDR receivers. The adjacent sub-bands, collectively spanning a transmit signal bandwidth were auto-correlated with a replica transmit signal to restore frequency and phase offsets. The offsets arise due to different local oscillator manufacturing tolerances, temperature effects and ageing. A 100 MHz bandwidth uniform white noise signal was reconstructed from both dual (2 fi 50 MHz) and multiple (4 fi 25 MHz) SDR collections. The 100 MHz bandwidth exceeds a B205 SDR receiver instantaneous bandwidth. The auto-correlation technique minimizes SDR hardware numbers as bandwidth overlap is not required. Hardware test Symbol Error Rate (SER) was compared with a theoretical coherently detected M-ary orthogonal signal. A 2 MHz dual SDR uniform white noise signal reconstruction exhibited a 5 dBW loss when compared with the theoretical value. The 4 MHz multiple SDR signal reconstruction exhibited a 6 dBW loss. Finally, a linear feedback shift register was used to generate the uniform white noise signal. This provided near true-noise characteristics employing a polynomial primitive to ensure 236 - 1 non-repeatable sequences.

AFIT Designator

AFIT-ENG-MS-20-M-049

Share

COinS