PROPAGATION PREDICTION AND PLANNING TOOLS FOR DIGITAL AND
ANALOG TERRESTRIAL BROADCASTING AND LAND MOBILE SERVICES
Satılmıs¸ Topcu, Hayrettin K¨oymen, Ayhan Altıntas¸, and ˙Irs¸adi Aksun
Communications and Spectrum Management Research Center (˙ISYAM)
Bilkent University, Bilkent, 06533 Ankara, Turkey
E-mail:
[email protected]
Abstract–The problem of efficient utilization of the radio-frequency spectrum is becoming more complex due to the increasing demands on the spectrum, which is a limited natural resource. If this natural resource is to be used efficiently, administrations need to use advanced Automated Spectrum Management Systems (ASMS) which would facilitate national spectrum man-agement and monitoring, coordination among adminis-trations and notification to International Telecommuni-cation Union (ITU). This paper presents some of the spec-trum engineering analysis and planning tools of the Na-tional Frequency Management System (NFMS) as part of the ASMS which has been developed for the admin-istration of Turkey. This system includes a Relational Database Management System (RDBMS) to support a database containing frequency assignment data, licence holder data and equipment characteristics data. The soft-ware is integrated with a Geographic Information System (GIS) and it has the capability of using digital terrain data and geographic data for advanced spectrum engi-neering calculations.
I. Introduction
Due to the growing demands on the radio-frequency spec-trum, there is a need to improve spectrum management tech-niques. The increase in the shared use of spectrum among administrations requires the use of more complex analy-sis methods. The efficient solution of spectrum manage-ment problems requires data storage and analysis capabili-ties, and consequently requires the application of computer-aided techniques for data management and frequency assign-ment. For this purpose, advanced Automated Spectrum Man-agement Systems need to be developed to effectively meet the requirements of spectrum management, to handle spec-trum management data and to add the capability of using digital terrain data for specific engineering calculations [1].
In this paper, we present some spectrum engineering tools used to assist in the assignment of frequencies and planning of digital and analog broadcasting and land mobile services. These tools are implemented as a part of the National Fre-quency Management System that has been developed for the administration of Turkey. Without loss of generality, we fo-cus on the analysis and planning of the terrestrial Digital Au-dio Broadcasting (T-DAB) and Digital Video Broadcadsting (DVB-T) services. In contrast to traditional analog
broad-casting, using a Single Frequency Network (SFN) of DAB or DVB systems one can distribute a broadcast program over all transmitters in the network within the same frequency block. In such a SFN, all the multiple delayed versions of signals ar-riving at the receiver are effectively combined yielding a net-work gain to obtain a received signal of significantly higher quality than in conventional single-transmitter systems. The interference may occur due to the signals coming from the transmitters in the same network but arriving later than the guard interval (self-interference) or due to the signals coming from other networks operating at the same frequency band (external interference).
The paper is organized as follows. Comparison of dif-ferent field strength prediction methods is presented in Sec-tion II. Analysis and planning tools for analog and digital broadcasting services are described in Sections III and IV, respectively, with some examples. Section V concludes the paper.
II. Propagation Prediction Methods
Propagation models are available in a wide range of com-plexity, accuracy, and input requirements. Some compar-isons between different prediction methods have been per-formed in previous studies [2, 3]. Various propagation mod-els recommended by ITU and some others given in the liter-ature with proven accuracy are implemented in this study for estimating the field strength, received power or transmission loss. The models employed here cover the frequency bands being currently used for both broadcasting services and land mobile services as well as other radio services, from VLF to UHF and higher frequencies. In order to allow predictions of the useful and interfering signals, calculations for differ-ent time and location percdiffer-entages can be done. Empirically gained attenuation correction terms are also added to take ac-count of the ground cover loss at the receiver site.
Propagation prediction schemes include ITU curves with h, clearance angle, mixed path corrections and other sug-gested modifications for VHF–UHF bands [4]. In addition, multiple diffraction is accounted for by following the ITU Recommendation 526 [5] and by Epstein-Peterson [6], Dey-gout [7] or Vogler [8] methods. For urban areas and higher frequencies, Okumura-Hata [9] and ITU Recommendation 1146 models [10] are implemented, respectively. For lower frequencies, standard programs developed and distributed by ITU are utilized.
Because of the complicated mountainous landscape of Turkey, a special study has been performed to compare the predictions of different methods when applied to real ter-rain. Fig. 1 shows a real terrain profile away from a trans-mitter which has an effective radiated power of 1 kW at f =30MHz. The terrain data is in the DTED Level 1 for-mat which has a resolution of 3 by 3 arcseconds.
0.0 10.0 20.0 30.0 40.0 Distance (km) 1.0 1.1 1.2 1.3 1.4 1.5
Elevation above sea-level (km)
Fig. 1. Terrain profile above sea-level from TX (N
39 30 0 30 00 , E32 37 0 45 00 ) in the180 azimuthal direction. Particular attention has been paid to various methods available for VHF and UHF bands. As seen in Fig. 2, ITU curves yield generally higher field strength predictions. Al-though inclusion ofhcorrection improves this, it is not to a satisfactory level because of neglecting the shadowing im-mediately after the hilltops, which can be accounted for by the inclusion of clearance angle correction. Multiple diffrac-tion models such as Epstein-Peterson, Deygout or Vogler are intended for the better prediction of diffraction loss. Shad-owing effect is similarly predicted by all of the diffraction models, but they differ more in the average signal level, as shown in Fig. 2. Note that Epstein-Peterson method predicts on the high side and the Vogler method does on the low side in terms of the average signal level. However, simplicity in applying the Epstein-Peterson method may make it more fa-vorable for some applications.
It is noted that the Vogler method does not specify how to chose significant knife-edges from the terrain elevation data. To solve this problem, an original selection procedure has been developed that accounts for both the distance and the depth of the valleys between local terrain maxima consid-ered as potential knife-edges. Since accounting for too many maxima grossly overestimate the losses, a flexible selection criterion has been introduced. The selection parametersis the fraction of the Fresnel zone used for making the decision whether two adjacent local maxima to be accounted either as two different edges or as a single dominating knife-edge. The decision is made by the rule that two maxima are
0.0 10.0 20.0 30.0 40.0 Distance (km) 0.0 20.0 40.0 60.0 80.0 100.0 120.0 140.0 |E| dB(uV/m) CCIR CCIR with Delta_h
CCIR with Clearance Angle Corr. CCIR with Epstein-Peterson CCIR with Vogler (s=0.6) CCIR with Deygout
Fig. 2. Comparison of the field strengths predicted by differ-ent corrections to ITU curves.
distinguished if there is a point in the valley located outside the fractionsof the Fresnel zone connecting the maxima. Alternatively, the maxima are not distinguished if the whole valley is located inside the given fraction of the Fresnel zone. The comparison of the field strength values using differents values is shown in Fig. 3. Note thats=0is the case when all the local peaks are accounted for as different knife-edges. As seen, this yields lowest values of the field strength. For practical purposes,s=0:6seems to be satisfactory.
0.0 10.0 20.0 30.0 40.0 Distance (km) -50.0 0.0 50.0 100.0 150.0 |E| dB(uV/m) Vogler, s=0.0 Vogler, s=0.6 Vogler, s=1.2
Fig. 3. Comparison of the field strengths predicted by the Vogler method with different selection parametersvalues.
As seen from the above discussions, large-scale field strength prediction models may yield substantially different results. The factors affecting the decision of which model to choose depend on the type of landscape, services and fre-quency band, but it may not be totally based on technical concerns. Some government regulatory bodies may prefer to impose one particular model.
III. Analog Broadcasting and Land Mobile
Services
The propagation simulations can be performed with any of the models described in Section II, and the type of simula-tion can be chosen as either coverage study or link study. Depending on the type of radio services, the study files gen-erated by simulation of the propagation model are processed to find coverage or interference areas, to calculate link avail-ability, to complete frequency planning and assignment pro-cedures, and to guide international coordination with neigh-boring countries. Any coverage or interference contour for a transmitter operating in broadcasting or land mobile service can be displayed on the map. One can directly determine the size of the coverage area or interference area and evalu-ate other useful data such as the population inside these re-gions. All these operations require the integration of various databases into the software. In addition, it is deemed to be necessary to integrate the software with a GIS to be able to display the simulation results together with the maps and any other spatial data such as roads, boundaries, etc. As an exam-ple, Fig. 4 shows the coverage area of an analog TV transmit-ter in ˙Izmir on the map background. The population covered by this TV station is calculated as 2082620. This simulation is carried out by using ITU curves and the effect of terrain is accounted for by using the reflection and multiple diffrac-tion technique in which Epstein-Peterson predicdiffrac-tion model is exploited.
Fig. 4. 70 dB(uV/m) coverage contour of Kaletepe, ˙Izmir (N 38 29 0 18 00 , E27 09 0 46 00
) analog TV station at UHF band with 40 dBW effective radiated power.
In the land mobile service, the coverage and interference areas are evaluated similarly as in the analog broadcasting service except that the receiving antenna height above ground level should be taken as 1.5 or 2 meters in the land mobile whereas it is taken as 10 meters in analog broadcasting. A study result for a land mobile base station operating at 150 MHz with 25 Watt ERP in Adana is given in Fig. 5. The
Fig. 5. -90 dBm coverage contour of Adana (N36 49 0 35 00 , E 35 37 0 53 00
) land mobile base station and the talk-back region of a mobile station operating in the same circuit.
coverage area is surrounded by -90 dBm contour drawn by a dark line. Throughout this area, the mobile unit can re-ceive the signals transmitted by the base unit. However, the base unit may not receive the signals coming from the mo-bile. Therefore, in order to determine a more realistic service area of a land mobile circuit consisting of one base unit and several mobile units, one should evaluate the talk-back range of the mobile. In the talk-back region shown as highlighted in Fig. 5, both the base and mobile units can communicate with each other.
IV. Digital Broadcasting Services
The software described here can calculate the useful (wanted) and interfering (unwanted) signal levels for a SFN as well as the network gain, protection ratio and coverage probability at all points in the study area defined by the user. It can also determine the coverage area or service area where the radio service with sufficient signal quality is provided while keeping the interference under a specified level. The service area is obtained by finding the regions where the cov-erage probability is higher than 95%for DVB and 99% for DAB.
A. Coverage Probability
The calculation of the coverage probability is split into three parts: calculation of the useful sum field strength, calcula-tion of the interfering sum field strength and evaluacalcula-tion of the coverage probability. For the first two parts, to perform the summation of wanted and unwanted field strengths, sev-eral approaches have been reported in the literature [13]. The location variation of a single field strength is modeled by a log-normal distribution with a standard deviation of 5.5 dB. For determining the power sums of log-normallly distributed stochastic variables, we use the LNM approach. The k-LNM approach is an approximation method for the statis-tical computation of the sum of distribution of several log-normally distributed variables [11, 12]. The method is based
on the assumption that the resulting sum distributions of the wanted and unwanted fields are also log-normal, the mean values and standard deviations of which are taken to be iden-tical with those of the true sum distribution. The k-LNM ap-proach is a modified version of the standard LNM obtained by introducing a correction factor to improve the accuracy in the high probability region. This procedure can be per-formed analytically. Suppose that there are givenn logarith-mic fieldsF
iwith Gaussian distribution (parameters F
i,
i, i=1:::n), i.e., the corresponding powers are log-normally distributed. The task is to determine the approximate log-normal distribution of the power sum, or equivalently, to find the parameters of the Gaussian distribution of the cor-responding logarithmic sum field as follows:
1. TransformF i,
i,
i=1:::n, from dB scale to Neper scale: F Neper = 1 10log 10 (e) F dB (1)
2. Evaluate the mean valuesM
iand the variances S 2 i of thenfields: M i =e Fi+ 2 i 2 i=1:::n (2) S 2 i =e 2Fi+ 2 i e 2 i 1 i=1:::n (3)
3. Determine the mean valueM and varianceS 2 i
of the sum field strength distribution:
M= n X i=1 M i (4) S 2 = n X i=1 S 2 i (5)
4. Determine the distribution parametersF
and
of
the approximate log-normal sum distribution:
F =ln(M) 2 2 (6) 2 =ln k S 2 M 2 +1 (7) wherekis a correction factor in the range0:::1. 5. TransformF
and
from Neper scale to dB scale:
F dB =10log 10 (e)F Neper (8) F and
are the mean value and the standard deviation of the Gaussian distribution of the logarithmic sum field. The k-LNM method suffers from the drawback that the correction factorkdepends on the number, the powers and the variances of the fields being summed. To obtain optimal results, an in-terpolation table would be neccessary, which is not suitable for an heuristic approach like k-LNM. For the sake of sim-plicity, we have chosen an average value ofk=0:7since the standard deviations of the individual fields are small.
Fig. 6. Useful sum-field strength levels in the service area.
Evaluation of the coverage probability is done by mul-tiplying the probability that the useful sum exceeds a spe-cific value with the probability that the difference between the useful and interfering sums exceeds a specific protection ratio. For determining the coverage probability, different ap-proaches are presented in [14, 15].
B. Examples
We have performed a case study of single frequency network (SFN) of digital video broadcast (DVB-T) in Konya lowland of Turkey which has a relatively flat terrain. The SFN consist of six stations located at the corner positions of a hexagon which has edges of 27 km long plus one station just at the center of the hexagon. The station at the center radiates 100 W and the remaining six stations on the edge of the hexagon cell radiate 1 kW each. The stations are separated by 27 km and their operating frequency is 826 MHz.
Fig. 6 shows the useful signal levels and the 48 dB(uV/m) contour that corresponds to the minimum required field strength for the SFN. The locations of the seven DVB trans-mitters are marked with a plus sign in the plot. Fig. 7 shows the network gain obtained by summing the wanted field strength values from individual transmitters. Note that the network gain becomes higher at the mid-points between two corner transmitters and it reaches the highest values at the points that are equi-distant to three transmitters. In Fig. 8, the coverage probability levels with the 95% coverage prob-ability contour are shown. In addition, the minimum field strength contour is also drawn in Fig. 8. It is noted that the 95% coverage probability contour is completely inside the minimum required field strength contour due to the fact that some transmitters behave as interferer in the region sur-rounded by these two contours.
Fig. 7. Network gain of SFN
Fig. 8. The required minimum field strength contour (outer) and 95% coverage probability contour (inner) on the cover-age probability levels for a 7-station DVB-T cell structure.
V. Conclusions
A propagation prediction and planning software tool for broadcast and land mobile services has been developed. The tool is integrated with various databases and a Geographic Information System. Through the use of real terrain data of Turkey, various propagation loss models are implemented and comparison results are reported. In addition, analysis and planning techniques for broadcasting, especially digital, and land mobile services are presented with some illustrative examples.
Acknowledgement
The authors wish to thank Supreme Council of Radio and TV of Turkey for their support for this project.
References
[1] ITU-R Recommendation SM.1370, “Design guidelines for developing advanced automated spectrum manage-ment systems (ASMS),” 1995.
[2] F.P. Fontan and J.M. Hernando, “Comparison of irregu-lar terrain propagation models for use in digital terrain databased radiocommunication system planning tools,”
IEEE Trans. Broadcasting, Vol. 41, no. 2, June 1995.
[3] M.V.S.N. Prasad and I. Ahmad, “Comparison of some path loss prediction methods with VHF/UHF measure-ments,” IEEE Trans. Broadcasting, Vol. 43, pp. 459-486, Dec 1997.
[4] ITU-R Recommendation P.370-7, “VHF and UHF propagation curves for the frequency range from 30 MHz to 1000 MHz,” 1995.
[5] ITU-R Recommendation P.526-5, “Propagation by diffraction,” 1997.
[6] J. Epstein and D.W. Peterson, “An experimental study of wave propagation at 850 MC,” Proc. Inst. Radio
Eng., Vol. 41, pp. 595-611, 1955.
[7] J. Deygout, “Correction factor for multiple knife-edge diffraction,” IEEE Trans. Antennas and Propagation, Vol. AP-39, pp. 1256-1258, Aug. 1991.
[8] L.E. Vogler, “An attenuation function for multi-ple knife-edge diffraction,” Radio Science, Vol. 17, pp. 1541-1546, Nov.-Dec. 1982.
[9] M. Hata, “Empirical formula for propagation loss in land mobile radio services,” IEEE Trans. Vehicular
Tech., Vol. VT-29, pp. 317-325, Aug. 1980.
[10] ITU-R Recommendation P.1146, “The prediction of field strength for land mobile and terrestrial broadcast-ing services in the frequency range from 1 to 3 GHz,” 1995.
[11] European Broadcasting Union, “Terrestrial digital tele-vision planning and implementation considerations,”
EBU Document, BPN005, Second Issue, July 1997.
[12] European Broadcasting Union, “Technical bases for T-DAB services network planning and compatibility with existing broadcasting services,” EBU Document, BPN003, Rev.1, May 1998.
[13] N.C. Beaulieu, A.A. Abu-Dayya, and P.J. MvLane, “Estimating the distribution of a sum of independent log-normal random variables,” IEEE Trans.
Communi-cations, Vol. 43, No. 12, Dec. 1995.
[14] H. Mokhtari, “Theoretical computation of the cover-age probability of a digital TV transmitter in an analog TV network using a stochastic approach,” IEEE Trans.
Broadcasting, Vol. 43, pp. 20-25, March 1997.
[15] A. Ligeti and J. Zander, “Minimal cost coverage plan-ning for single frequency networks,” IEEE Trans.
