MOBILE RADIO TRAN'SMISSION 1075 



especially for these tests and is similar to its foi'eninners which have 

 been used in the past for measuring- atmospheric static noise. Tlie LDR, 

 in combination with a caliliiated radio r(M'eiver, is capable of taking as 

 many as twenty instantaneous sami)les of radio sij>;nal str(Mif>;th per sec- 

 ond, sorting the samples by amplitude, and rendering infoiination on a 

 "batch" of samples from which a statistical distribution (•ur\"e can be 

 plotted. The LDR was also us(xl for measui'ing the statistical distribution 

 of audio noise in the output of the radio receiver. The LDR was, in this 

 case, associated with a special con\-erter whose characteristics resemble 

 those of a 2B noise measuring set. 



No arrangements were made for measuring radio proi)agation from 

 mobile unit to a land receiver. It was felt that the comparison by fre- 

 quencies woidd be substantially the same as in the outward direction 

 of transmission. It does not follow, however, that the background electri- 

 cal noise, against which an r-f signal must compete, will be the same at 

 mobile and land receivers. Strength of r-f signal required at land receiv- 

 ers for satisfactory transmission was measured at several typical locations. 



RECEIVED R-F SIGNAL STRENGTHS AND PATH LOSSES 



The first factor in evaluating mobile radio transmission is the strength 

 of the r-f signal which is received. This is inversely related to the loss in 

 the r-f path. The mobile units of a mobile system are either moving 

 around or, if stationary, are located at random. Since the effects of the 

 many geographical features, buildings, and the like, which influence 

 propagation can combine differently for different locations of a car, even 

 where the locations are only a fraction of a wavelength apart, the only 

 meaningful measure of signal strength is a statistical one. Such statistical 

 answers were obtained by making and recording many instantaneous 

 samples of field strength with the aid of the LDR, mentioned above. 



It is of interest to note that whenever the sample measurements were 

 confined to a relatively small area, say 500 to 1000 feet or less in extent, 

 the amplitude distribution of these samples tended strongly to follow 

 along the particular curve known as a Rayleigh distribution. Such a 

 curve and a typical set of experimental points are shown in Fig. 5. The 

 same distribution was obtained at all of the frequencies tested, including 

 3700 mc. The rapidity of signal fluctuation, as the car moved, was pro- 

 portional to frequency, but this does not affect amplitude distribution. 

 Such a distribution could have been predicted if it had been postulated 

 that the transmitted signal reached the car antenna by many paths 

 having a random loss and phase relationship. It is thus inferred that in 

 general the signal reaches a car by many simultaneous paths. 



