S Sw ARUP, V KUMAR & A AHMAD Himalayan Radio Propagation Unit, Dehra Dun Received 5 April 1975; revised received 4 November 1975

Transcription

1 Indian Journal of Radio & Space Physics Vol. 5, June 1976, pp Tropospheric Radiowave Propagation over Diffraction Paths* S Sw ARUP, V KUMAR & A AHMAD Himalayan Radio Propagation Unit, Dehra Dun Received 5 April 1975; revised received 4 November 1975 Measurements of field strength for a period of one year, were made over two paths in vhf band. One path (55 km) has two intervening ridges of Shivalik range and the other is a diffraction path (35 km) over curved earth. Distribution of hourly median values of one year data for both links are presented in the form of probability curves. Seasonal, monthly and diurnal variations offield strength have also been studied. The results of observations have been compared with theoretical and CCIR predictions. 1. Introduction In order to predict the performance of radio relay system in any particular geographical area, information about the long term median path loss and the characteristics of fading pattern are required. International Radio Consultative Committee (CCIR)l has given a set of curves for predicting the variability of transmission loss relative to the long term median for a number of climatic regions like continental, subtropical, etc. With a view to studying the validity of these curves for Indian conditions as well as for prep&ring such curves applicable to Indian climatic conditions, Himalayan Radio Propagation Unit (HRPU) has been executing a programme of field strength measurement for collection of radio propagation data. This paper presents th results of propagation studies conducted on two paths for one year, viz. Jan.-Dec. 71. The aim of the measurements was to collect long term propagation data over nonoptical paths and to assess the applicability of the existing methods such as CCIR curves; Atlas of Radiowave Propagation,2 Japan; and Bullington Nomograms3 for predicting the field strength/path loss for such paths in our country. 2. Details of Investigation 2.1 Sites Fig. 1 shows the general map of the triangular circuit covering Dehra Dun, Saharanpur and Roorkee. On account of the transmitting and receiving antenna heights, Roorkee-Saharanpur path is a beyond-the-horizon path of 35 km with intervening flat rolling terrain. Fig. 2 shows the profile of Roorkee-Dehra Dun path based on the effective earth's factor of 4/3. This is a double ridge diffraction p"lth of 55 km with the two ridges located *Paper Presented at the Symposium on Earth's Near Space Environment, ]8-21 February 1975, held at the National Physical Laboratory, New Delhi at distances of 36'0 and 47'2 km, respectively, from Roorkee. 2.2 Equipment Two transmitters operating at 73'2 and 82'8 MHz were installed at Roorkee. Field strengths of these transmissions were recorded at Saharanpur and Dehra Dun, respectively. The details regarding the transmitting and receiving systems are given in Table I. The two transmitters at Roorkee were so programmed that each was on for 20 min in an hour. Field strength samples of 20 min duration have been 3 o' 0' 2171 DISTANCE De.hTodun a \"'P 2872\2735 Mohan :.r '\29U \\ ; tf '., I(.../).,' 2g2 Milts t I +gale:-l't8mil<s 77130' 145'. 'i8' 10' - 7B1's' Fig. ]-Map showing the triangular circuit covering Dehra Dun, Sabaranpur and Roorkee + HEIGHT AeOVE m.l ROO'RKEE--265,2 m OEHRADUN-701-O m SAHARANPUR-271'3 m I III 1111 I II I ' 'I

2 SWARUP el al. : PROPAGATION STUDIES OVER DIFFRACTION PATHS RKEE ml HEIGHT-1m,2000 tt,60g'6m lst RIDGE 21'1d RIDGE 120m.. 750m DEHRADUN (HRPU) (701 0m) fearth RADIUS SCAlE:- DISTANCE-1 m,8 miles,12 87 km Pig. 2-Profile of Roorkee-Dehra Dun path Table I-Equipment Parametera for the Two Circuits Roorkee- Roorkee- Saharanpur Dehra DUD Frequency (MHz) 73'2 82'8 Path length (km) 35'2 55'2 Terminal elevations Continuous Horizontal wave 38'0 2'0 24'63 24'38 701'04 265'27 Dipole (m) above 271'26 265'17 Continuous Horizontal Dipole Vagi 4 10'97 2'0 (gain (a) transmitter mean sea site level for Receiving Transmi tter antenna antennaheight over a steady signal. This pattern has been observed during the entire period of recording. The amplitude of the ripple is found to be greater during the day hours vis-a-vis night hours. It is between 2'0 and 2'S,dB during 0800 to 1800 hrs and between 0'5 and I'S db during the remaining period of the day. Larger daytime variation in the amplitude of the signal is possibly due to the greater turbulence in the atmosphere during daytime. During night hours the atmosphere is relatively calm. Roorlree-Saharanpur path-a typical strip chart recording over this path is shown in Fig. 4. This' recorded signal may be broadly categorized into two types. It has been observed that during daytime the amplitude of the signal varies between 1'5 and 2'0 db at a rate of I cycle per 0'5 to I min. The nighttime signal amplitude varies between 3'0 and 4'0 db at a rate of 1 cycle per 2'5 to 3'0 min. During night hours the atmosphere is known to be stratified causing multipath propagation resulting in slow fading of the signal. The absence of slow fading during night hours in double ridge path of Roorkee-Dehra Dun may be due to the obstruction from the obstacle of the multipath component generated by the stratification of the lower atmosphere, 3.2 AnnuaJ Variation of Field Strengtb Statistical distribution functions of the recordings made on Roorkee-Dehra Dun and Roorkee Saharanpur paths are plotted in Fig. 5 as path loss w W 0: recorded for each hour of the day on 24 hr bas,is. The recorder charts were run at a speed of 12 in/hr. 2,3 Analysis The 20-min recordings were analyzed to find out the median value of the signal. These medians were considered as representative values for the hour since no fast fading of significant amplitude had been observed on either of the links. Further, on the average about 600 "twenty minute samples" were available for each month, which were utilized for drawing monthly and annual cumulative probability distribution of path loss. 3. Results 3.] Character of Obse"ed Recording Roorkee-Dehra Dun path-a typical strip chart recording over this path is shown in Fig, 3. The recorded signal appears like a ripple superimposed 32 o w UJ:> UJ-, 28 u> w«x UJ 36 Wu ;?; 32 in UI 28 m ọ.j TIME,hrs 1ST Fig. 3- Typical recording over Roorkee-Dehra Dun path 189

3 ' 15f ROORKEE -DEHRADUN L'NK G. - ROORKEE SAHARANPUR LINK \')', '0.'"Q tccir VALUES "0"0 INDIAN J. RADIO SPACE PHYS., VOL. 5, JUNE 1976 'o... I'll I'll. 7-5 '-0'" '-0'0, -o() 55 _ I PERCENT OF TOTAL RECORDING TIME THAT HOURLYMEDIAN e:xceeds DI-NAH: Fig. S-Distribution of hourly median path loss not exceeded against percentage of the total recording time. It is observed that, to a large extent, these curves follow a log normal distribution. However, there is departure for high values of path loss. It is also observed that for the longer path the curve has a slightly greater slope. The observed and computed values of path loss in excess of free space value for both circuits are given in Table 2. It is observed that the yearly median path loss in case of Roorkee-Dehra Dun circuit is very close to the one predicted by the Japanese atlas of radio wave propagation and is 6'5 db higher than the value predicted by the Bullington nomograms and 7'0 db more than the value calculated by formula given in the CCIR report CCIR-70. The difference between observed value and (CCIR) those predicted from last two methods Roorkee-Dehra Dun (exptl) -7'0 is due to the fact that the single knife-edge approximation as suggested in the two methods cannot be applied to this path which is essentially a double ridge path. The observed yearly median path loss for Roorkee-Saharanpur link is close to the values calculated from CCIR formulae (report 244-2) and Bullington nomograms. The long term variability of path loss is controlled by meteorological conditions of the area. For predicting the variability of path loss, CCIR have classified the world climate into eight types and prepared the variability curves for these climates. For comparing experimental results with the CCIR predictions, matching of the climate over the region with the CCIR climatic regions was carried out. It was found that in the northern plane of India annual mean of surface refractivity was 320 N units and annual range was 60 N units. This is closest to the continenta] subtropical (type 2) climate of CCIR classification having the corresponding values of 320 and N units, respective]y. The long term variation of path loss observed on the experimental links are compared in the Table 3 with the values obtained from CCIl{ curves for type 2 climate. Table 3 gives path loss observed for 1, 10,90 and 99% of the time. This gives an idea regarding the fade margin to be provided for achieving 90 and 99% propagation reliability as well as interference potential of such transmission for small percentage of time. It is seen from Table 3 that difference of observed values for 1, 10,90 and 99% level from 50% level for the two links is more than the corresponding CCIR values. In other words, a greater fade margin is to be pro vided for various percentages of propagation reliability than that predicted by CCIR curves. The interference potentia] of transmission for 1 and 10% level is also greater than the CCIR predictions..-- SO YEARLY MEDIAN- 28'j [G'O --- '8 5 --".'\.. "".,- '\, _,.,----_ -r-r- r-::f :.... Table 2-Path Loss in Excess of Free Space Value Radio path Observed yearly CCIR Computed value Atlas of radio median Bullington wave path nomograms Propaloss gation db (Japan) Roorkee- 28'S 21'S 22'0 29'0. Dehra Dun Roorkee 41"5 41'0 42"5 45"0 Saharan pur Table 3- Path Loss for Different Percentiles Relative to SO%Level Radio path Path loss (in db) for percentiles of 1% 10% 90% 99% Roorkee-Saharanpur +4' '5 +3' l's -2'S (exptl) -4 S ' '0 al "Q _J.I) o -' 20 W 30 o -- ROORKEE SAHARAN PUR t.ink ---- ROORKEE DEHRADUN LINK./. SO YEARLY <II g Jan. Feb. Apr, June Aug../. to SO!to _... Fig.6-Monthly variation of hourly median path loss (ordinate not to exceed for '10 of time shown on curves)

4 SWARUP It al. : PROPAGATION STUDIES OVER DIFFRACTION PATHS Table 4-Difference of 90% and Path Loss Value. Month Roorkee- Roorkee- Dehra Dun SaharanpUl' Jan, 6'5 6'0 Feb. Mar. Apr. May June July 8'S 18'S 12'5 6'5 11'0 4'0 6'0 70 7'0 10'0 9'0 Au,. Scpo Oct. 6'0 5'5 8'0 "5 :Nov. 4'0 Dec. 4'S 3.3 Monthl, Variations Roorkee-Dehra Dun /ink-the path losses observed for 10, 50 and 90% level for each calendar month are shown in Fig. 6. The median path loss of individual months fluctuates around yearly median value of 28'5 db within the limits of -3'0 and + db. The minimum median path loss is evident during January, and maximum median path loss during April. For April, 90 percentile of hourly median path loss deviates from yearly median path loss by a maximum of 10'5 db. The corresponding value for March is 1Q'O db and for all other months it is less than 10'0 db. The value of 10 percentile of hourly median path loss is minimum during March and it is 8'S db less than yearly median value. The differences between 90 and 10 percentile of path loss, for each month are listed in Table 4. These values give an idea of the extent of variation in hourly median path loss during a month. It is seen from Table 4 that maximum difference or scattering of the hourly median path loss was observed during March and minimum during July. Roorkee-Saharanpur link-fig. 6 illustrates 10, 50 and 90 percentile of hourly median path loss for different months. It is found from the curve of 50% level that monthly medians vary around yearly mean between -3'0 and +3'5 db limits. The value of path loss for 90% level is maximum during May and differs from yearly median value by 7.S db. The value of path loss for 10% level is minimum during October and is 6'0 db less than the yearly median path loss. The differences between path loss for 90 and 10% level are given in Table 4. It is seen from Table 4 that maximum scattering of median path loss occurred in May and minimum in July, and September. August 3.4 Seasoaal Variadoo Fig. 7 shows the statistical distribution of hourly median path loss during winter and summer months for the two links. For both links median loss has been found to be more during summer than 'in winter, For Roorkee-Dehra Dun link, the median path loss is 4'5 db more duriug summer than in winter, the corresponding value for the other link being 3'7 db. The path loss for 90% level is also more in summer than in winter for the two links and these values are 7'S and 4'5 db, respectively. For Roorkee-Dehra Dun link the probability distribution curve for summer months crosses over the probability distribution curve for winter months near 17% level and the two curves run close to each other thereafter. This indicates that high signal observed for smaller percentage of time during both the seasons is likely due to similar propagation mechanism. For Roorkee Saharanpur link also the two curves approach each other for lower percentile values, In general, the slope of distribution curve during summer is more than in winter for both links. This indicates greater variation of hourly median during summer than in winter, The difference in distribution of hourly medians path loss during two seasons is more pronounced for Roorkee- Dehra Dun link, 3.5 Diurnal Variation To study the diurnal variation of path loss, the daily data has been grouped into four intervals, i.e. from 00-06, 06-12, 12-18, hrs. The median values of path loss tor these intervals are listed in the Table 5 for the two links. It is observed from the Table 5 that the path loss, in general, is minimum dllring hrs and maximum during hrs; also the path loss during hrs is less than the value during hrs. It is evident that diurnal. variations are influenced by sun rise and sun set. The path loss increases slowly after sunrise, reaches a 20i en In g '" --IlOOIIKEE-llEHAAOUN link -----I1OORl<S-SAHAIlANI'UI1 l_ WlNTU NONTHS1No-'O fml 30 I 50 I 70 I eo 95 " PElleEM' OP " TH«T HOURLY MfOlAKORIlINATE Fig. 7-Seasonal variations of median path loss for the links two 191

5 ,I 30'8 46'6 28' '7 40'0 43'6 40'3 29'3 44'2 42'S 40'1 42'2 26'9 39'0 25'3 42'8 39'2 39'3 27'3 26'6 30'3 44'6 34'0 43'0 41'9 43'1 29'4 30'4 39'6 31'S 39'7 26'2 26,7 31'4 28'8 28'S 40'2 45'8 27'4 26'3 42'3 43'4 32'0 29'6 33'0 47'8 47'9 31'7 45'3 43'2 24'9 40'9 39'9 28'0 38'9 29'9 47'6 38' ' '7 41'8 46'1 45'1 42'7 38'6 44'1 39' '7 Table 24'8 26'6 25'8 28'4 32'7 28'1 26' '4 30'S 28'65-Diurnal INDIAN J, Variation RADIORoorkee-Saharanpur SPACE Path LossPHYS" in Excess VOL, of Free 5, JUNE Space1976 Value db Roorkee-Dehra Dun maximum in the afternoon, and decreases slowly after sunset attaining minimum at midnight. 4. Conclusions (i) Median path loss/field strength can be deduced quite accurately for curved earth diffraction path by means of formulae given in CCIR report or by the use of Bullington nomograms, (ii) CCIR report or single-knife edge method of estimating median path loss is not accurate for multiridge diffraction paths, More accurate theoretical methods are to be used for multiridge paths such as given in atlas of radio wave propagation (Japan), (iii) Greater fade margin, than calculated from CCIR curves, has to be provided for various percentage reliabilities, Also in planning the links, account should be taken of greater interference potential for low percentages, These results, therefore, suggest the need for extensive long term measurements in vhf/uhfjmicrowave ranges in our country, (iv) The recording of signal did not show any abnormally high signal. This indicates the absence of abnormal propagation conditions at this range of frequencies in this part. (v) For double ridge diffraction path, greater fade margin is required during summer than in winter months, (vi) The diurnal variation of path loss is influenc ed by the presence or absence of the sun, and a definite diurnal trend is marked, However, this variation is very small for such short paths and of little consequence, References 1. CCIR Rep, 244-2, Vol. 2 Study Group-S, New Delhi, 1970,87, 2, Atlas of Radio Wave Propagation Curves (The Radio Re;;earch Laboratories, Ministry of Postal Services, Tokyo, Japan), 3, Bullington K., Proc, IRE, 35 (1947), 1122, 192