Clark and Yarnall 



signal format is superior to straight cw in that the noise level is 

 continuously monitored (although, with the very narrow bandwidths 

 obtainable with the phase coherent demodulator (Fig. 3) , high noise 

 levels were rarely a problem). During the 15 seconds off period, 

 the phase trace will either present a random smear over 360°, or dis- 

 play a tendency to "hang up" at 180°. 



In Figure 3 the true relative angular positions of the source, 

 H3 and H43 have been indicated. The exaggerated vertical scale in 

 Figure 2 obviously creates a false impression. Figure 3 ignores the 

 multipath nature of the Straits of Florida propagation path, and the 

 complexity of the path boundaries. In this minimal form the exper- 

 iment exhibits the basic geometry of one of the simplest of the wave 

 transmission experiments--a study of forward scattering at essential- 

 ly zero scattering angles. A more complete physical interpretation 

 explicitly acknowledges the dominant influence of boundary reflec- 

 tions and treats the physics of the experiment with the concepts of 

 guided wave propagation. This paper, however, is intended to empha- 

 size techniques and data. For this purpose there is some advantage 

 in briefly considering the ideally simple case of fully coherent 

 forward scattering and discussing the results that would be obtained 

 with the experimental set-up of Figure 3. 



AN IDEAL EXPERIMENT 



We assume a cw disturbance to be transmitted from source to 

 receiver through a medium whose index of refraction is fluctuating 

 sinusoidally. The experimental geometry is such that all boundary 

 and interference effects are avoided, and the disturbance can be 

 regarded as a single "beam" of energy. The frequency of the sinus- 

 oidal change in the medium is further assumed to be very low compared 

 to the frequency of the propagation disturbance, and the wave length 

 of the medium fluctuation is of sufficient length that the whole of 

 the cw disturbance is affected. An extreme example might be the 

 case of a plane wave sound beam passing through a body of water which 

 is changing uniformly in temperature.* 



The result of this idealized experiment is a signal at the 

 receiver which is purely phase modulated at the medium fluctuation 

 rate. The Fourier spectrum, amplitude, and phase of the signal 

 might appear as in Figure 5. The phase modulation index, p, is a 

 function of signal wave length, propagation path length, and the 

 amplitude of the medium fluctuation. 



An investigator interested in studying the medium fluctua- 

 tion might choose to work directly with an "on-line" spectral ana- 

 lysis of the received signal. In this case the fluctuation frequency 



The analogous case of electromagnetic wave propagation through a 

 fluctuating plasma is discussed with mathematical detail by 

 Heald & Wharton (Ref. 5). 



311 



