194 BELL SYSTEM TECHNICAL JOURNAL 



Using this antenna for measurements in the vertical plane and another 

 identical antenna for measurements in the horizontal plane, angle-of-arrival 

 data were obtained during the summer of 1944 over a twenty-four mile, 

 partly over-water, path between Beer^s Hill, N. J. and New York and over 

 a thirteen mile over-land path between Beer's Hill, N. J. and Deal, N. J." 

 In the horizontal plane, deviations in the angle of arrival were rather un- 

 common and were not greater than ±0.1 degree from the true bearing of the 

 transmitter. In the vertical plane, angles of arrival above the true elevation 

 of the transmitter were observed to be as much as 0.5 degree on the New 

 York path and 0.3 degree on the Deal path during times of anomalous propa- 

 gation. From these measurements it was concluded that microwave re- 

 peater antennas could be made highly directive in the horizontal plane but 

 should have beam widths somewhat greater than J degree in the vertical 

 plane unless means for steering the beams are provided. 



Although the J degree beam width of these scanning antennas was sharp 

 enough to permit the separation of the direct and the water-reflected com- 

 ponents on the New York path, and to demonstrate the anomalous behavior 

 of each, there was evidence that occasionally there were signal components 

 so close together in angle that a sharper antenna would be required to resolve 

 them. Consequently a scanning antenna of the metal-lens type was con- 

 structed for operation at 1.25 centimeters. The aperture of this antenna 

 was 20 feet in the long dimension; the beam width was 0.12 degrees. Using 

 this antenna and also the 3.25 centimeter scanning antennas, angle-of- 

 arrival measurements were made in the summer of 1945 on the Deal-Beer's 

 Hill path. ^2 xhe most noteworthy result of these observations was the 

 demonstration of multiple-path transmission. Two, three and, at times, 

 four distinct signal components were observed simultaneously during one 

 night when the transmission was extremely disturbed. These transmission 

 paths generally were above the true direction of the transmitter; at one time, 

 a weak signal was arriving at an angle of 0.75 degree relative to the line of 

 sight. These components varied in angle of arrival and in signal amplitude. 

 Wave interference among them caused severe fading on broad beam antennas 

 that would accept all the wave paths. 



Another significant result of these angle-of-arrival measurements was 

 evidence that the transmission mechanism was very similar for wave- 

 lengths of 3.25 and 1.25 centimeters. Angles of arrival, measured simul- 

 taneously at the two wavelengths, agreed very well for times of single-path 

 transmission ; multiple-path transmission was observed on both wavelengths 

 although the 3.25 centimeter antenna was too broad to resolve the com- 



" W. M. Sharpless, "Measurement of the Angle of Arrival of Microwaves", Froc. 

 I. R. E., vol. .34, pp. 837-845; November 1946. 



" A. B. Crawford and W. M. Sharpless, "Further Observations of the Angle of Arrival 

 of Microwaves", Proc. /. R. E., vol. 34, pp. 845-848; November, 1946. 



