Date of Award

5-2009

Document Type

Thesis

Degree Name

Master of Science (MS)

Legacy Department

Computer Engineering

Advisor

Pursley, Michael

Committee Member

Noneaker , Daniel

Committee Member

Russell , Harlan

Abstract

Receiver-based statistics can contain useful information about the quality a wireless channel. We examine methods of generating receiver statistics without real-time simulation of the receiver and the decoder, particularly during a network-level simulation. Attention is restricted to a receiver employing binary orthogonal signaling, noncoherent demodulation, and a soft-decision decoder in a frequency-hop network. Data is gathered from off-line simulations of channel, demodulation, and decoding processes, using a family of turbo product codes. Study and analysis of the data are used to characterize the behavior of the receiver statistics and determine on-line methods of generating the statistics in network-level simulations. To measure the performance of our generation techniques, we compare the results with simulation data.
The goal of the generation methods is to provide receiver-based statistics for simulations of higher layer protocols without the need for simulation of the decoding process, which is time consuming. With this goal in mind, we design our generation methods to be simple and fast, favoring approximations over exact replications whenever the resulting simplification is significant. Both demodulator and decoder statistics are considered. Results are given which demonstrate the fidelity of our generation methods.

Share

COinS