Page 230 - ITU Kaleidoscope 2016
P. 230

2016 ITU Kaleidoscope Academic Conference




                  5. GROUND TRUTH EVALUATION

          As argued in the paper that there are no propagation mod-
          els clearly known to perform better in regions of the devel-
          oping world that can be used in geo-location databases, the
          first step that countries in these regions should take is to per-
          form extensive spectral measurements and compare values
          of the path losses obtained from the measurements against
          those estimated by propagation models suggested for use in
          geo-location databases. Therefore, the authors did a limited
          physical evaluation of the approaches by conducting spec-
          tral measurements and comparing values of the path losses
          obtained from the measurements against those estimated by
          some common propagation models. The experiments were
          conducted in the city of Cape Town in South Africa. This
          section gives a detailed discussion of how the whole process
          was carried out.


          5.1. Propagation models examined                      Figure 3: Measurement sites relative to BS transmitter

          Five propagation models were examined and compared with
          values from measurement data; Longley-Rice (Irregular Ter-  a Nagoya NA-773 wide band telescopic antenna with verti-
          rain Model), Hata for urban areas, Egli, Ericsson 9999 and  cal polarization, which has wide band measurement capabil-
          Free Space Path Loss (FSPL).                       ity. The RF Explorer was connected to an Android phone
                                                             installed with an android code that starts to measure spec-
          5.2. TV transmitter used                           trum immediately after the RF Explorer is connected using
                                                             an On-The-Go (OTG) cable. At each site, spectrum monitor-
          An analog terrestrial television (ATT) transmitter of one of  ing was done for 8 hours from 08:00 in the morning to 16:00
          the public TV broadcasters in South Africa called South  in the afternoon.
          Africa Broadcasting Corporation 2 (SABC2), located on
                                               00
                                             0
                                          ◦
                         00
                      0
                   ◦
          latitude 33 52 31 S and longitude 18 35 44 E, was used
                                                             5.5. Results
          as a base station (BS) transmitter. Its transmission param-
          eters obtained from the Terrestrial Broadcasting Frequency
                                                             The 3-hour measurements at each measurement location in
          Plan 2013 document by the Independent Communications
                                                             channel 22 from which the BS transmitter was broadcasting
          of South Africa (ICASA) [10] are: UHF channel = 22, fre-
                                                             were averaged and the mean value was taken as the received
          quency = 479.25 MHz, Effective Radiated Power (ERP) = 2
                                                             signal power R x . As a starting point of our analysis, we de-
          KW = 63.01 dBm, antenna polarisation = vertical.
                                                             cided to confirm the square law dependence of power loss
                                                             with distance first. We used the average value of the mea-
          5.3. Measurements points                           sured power at the closest point to the BS transmitter (SITE
                                                             1) and estimate from there the power received at longer dis-
          Twelve locations located at different distances from the BS  tances along the same approximate radial by using the Friis
          transmitter site were identified where measurements were  transmission equation [12] (equation 1).
          done. Tables 1 shows the geographical positioning system
          (GPS) coordinates of the sites and their distances away from  Pr(d) = Pr(d o ) + 20 ∗ log(d o /d)  (1)
          the BS transmitter. The table also shows values of the height
          above avarage terrain (HAAT) of the sites calculated using  where Pr(d) is the received power at distance d in
          GLOBE 1 km Base Elevation database [11] with the number  the same radial where d o is calculated, Pr(d o ) is the
                                        ◦
          of evenly spaced radials equal to 360 in each case. Figure 3  received power at a close-in-reference-distance d o .
          shows the measurement points relative to the BS transmitter
          generated using google maps.                       In that way, we were able to compare the measurements with
                                                             those values obtained using equation 1. As Table 2 shows,
                                                             there is reasonable agreement between the measured and cal-
          5.4. Spectrum measurements setup
                                                             culated values. This confirm that the square law dependence
          Outdoor spectrum measurements in the UHF ATT frequency  of power loss with distance is adequate.
          band were done at the locations using a hand-held RF Ex-  The accuracy of an Effective Isotropic Radiated Power
          plorer model WSUB1G, which has a measurement frequency  (EIRP) of a transmitter, the gain and return loss of a receiv-
          range of 240 MHz to 960 MHz. The model was fitted with  ing antenna determines the real path loss. In this experiment,



                                                          – 212 –
   225   226   227   228   229   230   231   232   233   234   235