Abstract

This these explores automatic modulation recognition as applied to PCM/FM, SOQPSK- TG and ARTM CPM. It found that the likelihood based approach is intractable. The statistical features of the amplitude, phase and frequency are ineffective at distinguishing these modulation types. A method based on the phase changes between symbols is developed and shows that as long as symbol timing is established, this method can effectively distinguish PCM/FM, SOQPSK-TG and ARTM CPM for signal-to-noise ratios above 30 dB. Another method, the Bianchi-Loubaton- Sirven technique, was able to distinguish PCM/FM and SOQPSK-TG but was unable to distinguish ARTM CPM. A happy byproduct of this classification algorithm is a reasonably accurate estimate of the bit rate. Simulation results show that this classifier works essentially error-free for signal- to-noise ratios above 20 dB and for sufficiently high resolution in the search algorithms required by the maximizations.

Degree

College and Department

Ira A. Fulton College of Engineering and Technology; Electrical and Computer Engineering

Rights

http://lib.byu.edu/about/copyright/