3 . ESTIMATED SYSTEM ACCURACIES 



The estimated accuracies achievable under the most favorable operat- 

 ing conditions for the common types of measurement methods are illustrated in 

 Figure III-IO. In this figure the curve on the left is calculated on the assump- 

 tion that pulse comparison measurements may be made to about one -fifth of the 

 pulse length . Since it is difficult to radiate a pulse whose length is less than 

 50 cycles of the carrier frequency, time -difference measurements between 

 pulses can only be made at about one -fifth the pulse length . The measurement 

 is therefore accurate to about 10 wave lengths in space . If we assume that two 

 measurements are made and two lines of position which are at right angles to 

 each other are established, the average error of fix will be about ^2 times the 

 least reading . This leads to the fix error shown . 



If the two pulses to be compared are carefully made similar when 

 radiated and are amplitude -equalized at the receiver, then visual superposition 

 is accurate to within about 1% of pulse length . This makes possible a 20 -fold 

 improvement in the accuracy as shown by the center curve in the figure . 



So far as is known, the final step in accuracy improvement is phase 

 comparison, accurate to about 1-3 degrees of phase . Taking 1/200 wave length 

 as an average figure, this represents an accuracy 100 times that of Loran-A, 

 as the curve on the right in Figure III -10 indicates . 



REFERENCES 



J. A. Pierce, "Electronic Aids to Navigation, " in Advances in Electronics , 

 Academic Press, 1948, pp. 425-451. 



28 



arthtir Zn.Htttlc.Knr. 



