OPTIMIZING THE LOW PROBABILITY OF INTERCEPT PERFORMANCE OF A SPREAD SPECTRUM SYSTEM (CODING, BOUNDS, COMMUNICATIONS)
Abstract
This thesis investigates two problems associated with the design of a low probability of intercept spread spectrum system: Finding a general method for measuring the low probability of intercept performance of such a system, and evaluating the performance improvements due to the use of various power reduction methods within such a system. Two methods for measuring the low probability of intercept performance are proposed and described: An interception-time figure of merit and a probability of intercept figure of merit. The latter of the two was favored and studied in detail. The performances of direct sequence and frequency-hopped spread spectrum systems are evaluated, using the chosen figure of merit, as functions of both constrained and adjustable system parameters, for various levels of interceptor a priori knowledge. The energy detector and several envelope detectors are studied in detail, and are considered the available building blocks in the interception systems studied. The detection performance of each of three envelope detector systems is analyzed for each of two signals: A time-limited sinusoid and a time-limited FSK modulated carrier. The direct sequence signal is found to be more difficult to detect than the frequency-hopped signal, in general. Comments are made regarding situations where the frequency-hopped signal may be more difficult to detect. Design guidelines are specified for determining frequency-hopped system parameters that optimize the low probability of intercept performance. Partial-band detection of direct sequence signals is proposed, analyzed, and found to offer a measurable detection performance improvement. The most successful power reduction technique studied is the use of error correction codes. Lower and upper bounds on the required receiver signal-to-noise ratio are determined as functions of specified system parameters. These bounds imply upper and lower bounds on the power reduction realized from coding. The effect of a specified burst interference environment on the upper bound for signal-to-noise ratio is also determined. Multiple carriers and expanded signaling alphabets are analyzed to determine potential performance improvements.
Degree
Ph.D.
Subject Area
Electrical engineering
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.