Design and analysis of multiple frequency signal sets used in severe multipath environments

James Douglas Callison, Purdue University

Abstract

This thesis first concerns itself with eliminating fading effects associated with the two path channel. This study has attempted to use worse case channel models such that the design of the multiple frequency signal sets will be more robust than their previous counterparts. By transmitting several optimally chosen frequency components simultaneously, a large improvement in received signal power is achieved over a specified range of multipath delays. It is also shown that the number of frequencies and the required frequency spacing are reasonably small. The problem of radar range tracking is then considered, where signal-to-noise ratio is the main criterion. It is shown that with a multiple frequency signal set the tracking system no longer has to coast through regions where the reflection almost completely cancels the direct signal. An important advantage of this system is that the target does not have to be required everytime it comes out of one of the multipath nulls. Also, the study of a channel with many reflected paths is undertaken. In this research the concept of data communication in a factory environment is considered. Unlike the radar case, the factory environment is not as concerned with the SNR problem as it is with the probability of detection of the signal. For this case the optimal worst case multiple frequency signal sets are computed for a noncoherent detector. Also, the chirp signal is shown to be a natural diversity type signal set which uses only one carrier transmission, unlike its multiple transmitted counterpart. (Abstract shortened with permission of author.)

Degree

Ph.D.

Advisors

McGillem, Purdue University.

Subject Area

Electrical engineering

COinS