4-8] AMPLITUDE, ANGLE, AND RANGE NOISE 207 



the target is 0.134. For equal reflection centers uniformly spaced over a 

 circular area, this fraction becomes 0.2. 



Data on angle noise have been collected at the U. S. Naval Research 

 Laboratory in connection with investigations of tracking noise. This work 

 has been done at a frequency of about 9400 Mc, so that it is applicable to 

 airborne radar problems. Examples of angle fluctuations of the type 

 discussed above are shown for an R4D in Fig. 4-28. The variations in 

 apparent reflection center are seen to be greater than the linear dimensions 

 of the aircraft. Also, the deviations at the 90° aspect are larger than at 0° 

 and 180°. 



Fig. 4-29 shows the spectrum^" of angle noise for several runs of an SNB. 

 The angular noise in this spectrum has been multiplied by the target range 

 so that the spectral density is independent of range and expressed in yards 

 per Vcps- The amplitude decreases fairly regularly with increasing fre- 

 quency, so that the total noise power is finite. In many cases the spectrum 

 can be fitted satisfactorily by a curve of the form 



A = A,{\ ^r/U)-'i'~ (4-58) 



which corresponds to the transfer characteristic of a single-section RC 

 low-pass filter. 



Since the vertical span of an aircraft is much less than its horizontal span, 

 angle noise of a single aircraft is much less in elevation than in azimuth. 

 In low-angle tracking, however, reflection from the ground or sea has the 

 effect of creating an image aircraft at an equal distance below the surface 

 (see Fig. 4-13). If the angle between the target and its image is less than 

 the elevation beamwidth, the two will not be resolved. Variations in the 

 phase difference between direct and reflected rays then will cause the 

 effective reflection center to wander between target and image or beyond 

 them. This has been observed to be the case. As the range decreases, the 

 angular fluctuations increase until target and image can be resolved and the 

 tracking system locks on. However, it is possible for lock-on to occur on the 

 image instead of the target! 



Multiple targets have a similar effect on azimuthal variations. Thus, 

 multiple targets which are not resolved give rise to a much higher level of 

 tracking noise than a single target. 



Range Noise. In addition to causing angle noise, fluctuations of the 

 effective center of reflection of the target can give rise to fluctuations in 

 range, or range noise. Fig. 4-30 shows typical time plots of range noise for 

 several classes of target. Fig. 4-31 shows the range noise of a single SNB at 



^OThe observation time included in this spectrum is about 80 seconds, so that spectral 

 frequencies below about 3^ cps are cut off by analyzer limitations. 



