REGIONAL SFN DESIGN FOR C2G VARIANT DVB-T SERVICE AT FIXED ANTENNA RECEPTION MODE IN JABODETABEK AREA

Transcription

1 1 REGIONAL SFN DESIGN FOR C2G VARIANT DVB-T SERVICE AT FIXED ANTENNA RECEPTION MODE IN JABODETABEK AREA Tito Ilyasa 1,Dadang Gunawan 2, Wireless and Signal Processing (WASP) Research Group, Electrical Engineering, Faculty of Engineering, University of Indonesia, 1 fland_wmgc@yahoo.com, 2 guna@eng.ui.ac.id Abstract A lot of problems that found in analog broadcasting necessitate Ditjen Postel as the telecommunication regulator in Indonesia to promptly decides the time to migrate from analog to digital broadcasting. Digital broadcasting standard for television that will be implemented in Indonesia is DVB-T. When planning a network for DVB-T service, SFN method can be implemented. The advantanges of using DVB-T as digital broadcasting are more efficient bandwidth, increased robustness from noise, and lower energy consumption. Using SFN for network planning also directs to higher spectrum effieciency and location probability, as well as network gain that leads to power efficency. The purpose of this paper is to get the regional SFN specification for DVB-T sevice with optimal coverage. To achieve that, the transmitters need to be synchronized,including time delay and ERP parameters variation, and fill-in Tx as repeater addition. The designing aspects which are considered include modulation scheme, code rate, carrier mode, guard interval, frequency channel, bandwidth, reception mode, coverage probability, and network reference. The design is executed using the Chirplus_BC software at Ditjen Postel. Tx placement done based on network reference, while the parameters are adjusted according to designing aspects. After synchronization is done, the result fits network specification for optimal coverage and interference-free. Index Terms SFN designing, DVB-T service, fixed antenna reception. T I. INTRODUCTION HE expansion of broadcasting technology usage and the increasing of citizen s need expose the disadvantages of the analog transmission. For example, from the signal quality point of view, along with the development of the telecommunication technology, the increasing number of signals on air causes analog broadcasting signal easier to interference. Besides, with the increasing of the citizen s need, it lessens the availability of frequency spectrum. These adverse the analog broadcasting since one frequency channel only fits one service. Therefore, in 1990, the researchers from various global telecommunication organizations have begun to develop TV broadcasting using digital transmission. The advantages of using the digital transmission instead of analog transmission are [1]: More efficient in bandwidth usage using compression technique, which the digital broadcasting able to fit 2-10 TV programs in one frequency channel. Better signal quality because digital signal is more persistent from noises with OFDM technique and error correction. Lower energy consumption compared to analog television at same reception mode. Possibility of broadcasting one signal with different audio, video, and data services. Able to being used for various reception modes, like fixed antenna, portable, and mobile. Digital TV network design using the SFN can retrenches the usage of frequency spectrum. The broadcasting migration from analog to digital is the alternative chosen by Ditjen Postel. The DVB-T standard from Europe is adopted to be implemented in Indonesia. During migration period, Ditjen Postel needs to make a proper planning for network and frequency distribution for digital TV that will be implemented in Indonesia. The network design can uses either MFN or SFN method. II. DIGITAL VIDEO BROADCASTING WITH TERRESTRIAL TRANSMISSION A. MPEG-2 DVB adopts the MPEG-2 compression technique as data containers for audio and video source coding, system information, and multiplexing. Digital compression technique can being used to produce different bit rate that accustoms QoS requirements. On the other hand, digital TV technology also enable to combine several MPEG-2 s encoded to broadcast several digital TV services in same frequency channel. B. Coded Orthogonal Frequency Division Multiplexing DVB-T system broadcasts the MPEG-2 encoded TV signal with terrestrial transmission. Therefore, it needs suitable adaptation from digital code streams into terrestrial channel with different characteristics. This reason causes the usage of multicarrier modulation called COFDM technique, combined

2 2 with error correction. COFDM is generated with IFFT technique in the modulator and FFT technique in the demodulator. The technique is very robust in multipath reception and very useful for channels with linear distortion. COFDM uses thousands separated carriers to carry data signal, where the data is divided for each carrier through time division multiplexing process. COFDM technique also can use the different number of carriers, modulation scheme, and guard interval to maximize spectrum efficiency thus can operated in SFN. 1) Carrier Mode In DVB-T specification for COFDM implementation, there are two usable carrier mode with different number of carriers, that is 2k mode with 2048 point FFT and 8k mode with 8192 point FFT[2]. 2k and 8k mode implementation comparison shown by following Table I. TABLE I 2K AND 8K MODE IMPLEMENTATION COMPARISON [1] 8k Narrower carrier space Greater number of carriers Longer symbol period and guard interval For large area network and fixed antenna reception 2k Wider carrier space Less number of carriers Shorter symbol period and guard interval For small network and portable antenna reception 2) Guard Interval When a modulation applied to the carriers in OFDM, then the time period will be divided into active symbol period and guard interval period. Guard interval defined as time period within total symbol period when no new data is modulated onto carriers. Guard interval allows the reception in multipath environment, where time delayed signals are added to form complete reception signal. During guard interval period, any time delayed signals that is received will added to main signal without causing interference. C. Modulation Scheme COFDM can use QAM or QPSK modulation technique to separately modulates each carrier. The network designer can determines which modulation scheme is used to modulates signal partion to each carrier. QPSK will map 2 bits per symbol onto carrier, 16-QAM map 4 bits, and 64-QAM map 6 bits[2]. This mapping process allows constellation forming in phase space plane with gray mapping as shown in Figure 1. Fig. 1. QPSK, 16-QAM, and 64-QAM modulation constellation [3] Table II shows the implementation comparison of those modulation types. TABLE II QPSK, 16-QAM, AND 64-QAM MODULATION IMPLEMENTATION COMPARISON [1] QPSK 16-QAM 64-QAM Low bit rate Medium bit High bit rate Low minimum field strength rate Medium minimum field strength High minimum field strength D. Inner Coding Inner coding that used in DVB-T system is Rate Compatible Punctured Convolutional (RCPC) code[2]. In convolutional code, one or several information bits that become input will produce output with one additional bit as protection bit. For example, if we input one bit, then the ouput will be two bit. This means it uses coding rate 1/2. Coding rate can be defined as ratio of information bits to total bits. In DVB-T system, there are five eligible code rates, those are 1/2, 2/3, 3/4, 5/6, 7/8. When designing a network, we can select the different code rate to adjusted with the service error correction level. The higher error protection causes the decreasing of data rate. Code rate comparison shown in Figure 2. Fig. 2. Code rate comparison [1] III. SFN DESIGNING The ability of COFDM to prevent multipath interference, enable DVB-T technology to distribute program to entire

3 3 transmitters inside a network using same frequency. Here the Single Frequency Network (SFN) concept comes, where all transmitters inside a network operate in same frequency channel, modulated synchronously with same data signal, also transmitted at the same time. In SFN, the received signal is the superposition from all the signals that come from every transmitter inside the network which distribute the required program. The example of SFN model is shown in Figure 3. Bandwidth selection based on the bandwidth used in existing analog broadcasting, which is 8 MHz. According to the allocation adjusted by Ditjen Postel, digital broadcasting places the VHF and UHF frequency band. The VHF band ( MHz) is allocated for Digital Audio Broadcasting (DAB) service, while DVB places UHF band ( MHz)[5]. DVB-T channel allocation is shown in Figure 4. Fig. 3. SFN model example [4] This basic principle of SFN depends on proper synchronization of transmitters in the network. Synchronization is done by distributing stable frequency source and timing reference pulse to all main stations. Also, the important thing is all transmissions must have identical bits to the original data signal. This adjustment can be done by accurately chooses the system mode (2k or 8k) and varies the proper guard interval duration for transmission delay required by the network. A. Designing Aspects 1) DVB-T Variant DVB-T technology can use various modulation techniques, code rate, and guard interval. The utilization of one of the modulation technique can be combined with different code rate and guard interval, vice versa. Therefore, in its implementation, DVB-T system has many variants depends on combination of modulation technique, code rate, and guard interval that used. The code for each variant is shown in following Table III. TABLE III DVB-T VARIANT CODES[1][2] Fig. 4. UHF band allocation for DVB-T with SFN From 42 frequency channel in UHF band, channels ( MHz) allocated for DVB-T with SFN[5]. The channel that used in this designing is channel 51 (714 MHz). 3) Reception Mode and Coverage Probability In DVB-T implementation for broadcasting service, coverage area analysis can be done for various antenna reception mode, such as fixed, portable indoor, portable outdoor, and mobile. The reception mode that used in the design is fixed antenna reception mode. Fixed antenna reception mode can be defined as reception mode where antenna representation in reception area assumed as directional antenna placed on roof level. When calculates the equivalent field strength, the reception antenna height in this mode is assumed 10 m[3]. 4) Network References The selection of network reference is based on the SFN type consideration. This design uses the regional SFN type which includes Jabodetabek area. With the consideration of area dimension, various earth surface relief, and area type, then the suit network type for this designing is SFN network for small service area with dense transmitter placement. This network type consists of six hexagonal form, where each hexagon has 3 transmitters that placed in 3 different edge forming the triangle. The individual hexagon is shown as following Figure 5. From the ITU recommendation and designing type considerations, the DVB-T variant that used in this designing is variance with 64-QAM modulation type, code rate 2/3, 8k carrier mode, and 112 µs guard interval, or in other word C2G variant DVB-T. 2) Bandwidth and Frequency Channel

4 4 Fig. 5. Network Reference for regional SFN with small service area[6] The parameter specification for this network type is shown in Table IV. Fig. 6. Flowchart of Tx placement using Chirplus_BC software TABLE IV NETWORK REFERENCE PARAMETER FOR REGIONAL SFN WITH SMALL SERVICE AREA [6] To determine the power budget of this network type, it used the ERP from reference frequency 650 MHz that valued 31,8 dbw can be implemented in equation (1)[7]: ( ERP) ( ERP) D D = R + 20log (1) f R Annotation: (ERP) D = desired ERP value (dbw) (ERP) R = reference ERP value (dbw) f D = desired frequency value (MHz) f R = reference frequency value (MHZ) This design uses the frequency channel 51 (714 MHz), therefore from equation (1) the resulting desired ERP is 32,62 dbw. B. Transmitter Placement This section describes the step to determine transmitter placement in Jabodetabek area using Chirplus_BC software. The step sequence is shown in following Figure 6. f Fig. 7. Tx placement based on network reference Figure 7 shows the Tx placement in Jabodetabek area based on network reference. After placing the Tx in software, then the parameter of each Tx is adjusted referring to following Table V. TABLE V INITIAL TX PARAMETER

5 5 before. So, to obtain the optimal coverage that covers the entire service area, it is necessary to synchronize the transmitters in the network. B. Transmitter Parameter Synchronization 1) Time Delay Variation The addition of the time delay in a transmitter is done only when self interference occurred even the inter-transmitter distance doesn t exceed the DVB-T guard interval maximum distance. Self interference in this case occurred because the relief of the earth surface conduce several of the multipath signal having the arrival time longer than it has to be, thus they arrive outside the guard interval and causing interference. The determination of the time delay variation is done by previously calculates the maximum time delay with subtracts the value of transmission delay to value of guard interval that used. After attempts different variation of time delay to the transmitters with potential self interference, the results of the optimal variation of time delay is obtained and shown in Table VI. IV. SFN SYNCHRONIZATION ANALYSIS A. Coverage Calculation First, the value 52,545 dbμv/m of the minimum field strength is inputted into Chirplus_BC software for coverage calculation and analysis. The other parameters are adjusted from the designing aspects. TABLE VI OPTIMAL TIME DELAY VARIATION Tx Name Time delay (μs) TX_Tangerang 1 20 TX_Tangerang 2 0 TX_Tangerang 3 0 TX_Jakarta 20 TX_Depok 0 TX_Bekasi 1 20 TX_Bekasi 2 20 TX_Bogor 3 0 TX_Bogor 4 0 TX_Bogor 2 0 TX_Bogor 1 20 From the time delay variation above, the new SFN coverage is obtained, as shown by following Figure 9. Fig. 8. Coverage awal SFN From the Figure 8 above, it is shown that the area covered with green color is the area covered by SFN. Areas in red circle show that there are areas in Jabodetabek that haven t been covered by DVB-T service. The cause is self interference that occurred when multipath signals arrive in receiver outside the DVB-T system guard interval that has been adjusted

6 6 Fig. 9. SFN coverage with time delay variation In Figure 9 above, it is seen that the coverage result is better than before. But, there are still a few areas that haven t been covered by SFN. So, it is necessary to vary other parameter to obtain optimal coverage. 2) ERP Variation The next parameter to varied is ERP. The value of the ERP really influence the coverage, because the bigger value of the ERP means the signal can received well for longer distance. Different from time delay variation which based on the maximum time delay, the ERP variation is done based on the initial value of ERP. By combining several variation of ERP, the optimal variation of ERP is obtained and shown in following Table VII. TABLE VII OPTIMAL ERP VARIATION Tx Name ERP (dbw) TX_Tangerang 1 32,62 TX_Tangerang 2 32,62 TX_Tangerang 3 32,62 TX_Jakarta 32,62 TX_Depok 32,62 TX_Bekasi 1 40 TX_Bekasi 2 35 TX_Bogor 3 35 TX_Bogor 4 32,62 TX_Bogor 2 28 TX_Bogor 1 35 By recalculates the coverage for this ERP variation, the new SFN coverage is obtained as shown in Figure 10. Fig. 10. SFN coverage with ERP variation According to Figure 10 above, several areas which haven t been covered before, now have been covered. The recent coverage is nearly optimal, only small part in the south-east and north-west of Kabupaten Bogor that not covered. C. Tx Addition as Reepater Tx addition as repeater is done only if the synchronization of Tx parameters hasn t produce the optimal coverage. The placement of the Tx as repeater in the software done by estimating the center of the mountain area that haven t been covered. Repeater is not function as the main Tx, therefore the value of antenna height and the ERP that adjusted are not as big as the main Tx. Other parameter like time delay has to be adjusted too, because although it is function as a repeater, it is still potential for self interference. The remaining parameters are same with the main Tx. After adjustment done, the optimal parameter for the repeater is shown in Table VII. TABLE VIII REPEATER PARAMETER Next, the recalculation is done, so that the optimal SFN coverage for the Jabodetabek area is obtained, as shown in Figure 11.

7 7 [2] O Leary, S., Understanding Digital Terrestrial Broadcasting, Digital Audio and Video Series, pp.7, 61-63, , , , 118, 168, Artech House, Inc., Norwood, [3], Implementation Guidelines for DVB-T Transmission Aspects, in DVB Document A037, pp.9, 35, 38, 42, DVB Project Office, [4] Ligeti, A., Single Frequency Network Planning, Dissertation, pp.4-5, Royal Institute of Technology, [5], Konsep Draft Peraturan Perencanaan Frekuensi Untuk Keperluan Penyiaran, pp.3, 15, 20, Ditjen Postel, [6], Technical Criteria of DVB-T and T-DAB Allotment Planning, in ECC Report 49, pp.28, Electronic Communication Committee, Copenhagen, [7], Requirements for DTT Service Operating in the Frequency Bands MHz and MHz, in MCMC SRSP-521 DTT, Appendix J, Suruhanjaya Komunikasi dan Multimedia Malaysia, Selangor Darul Ehsan, Fig. 11. Optimal SFN coverage V. CONCLUSION From the initial value that applied to every transmitter, the optimal coverage hasn t been obtained yet. Therefore, it is necessary to do synchronization. Thus, the following new specification is obtained : The transmitters that get additional time delay from the 0 μs initial value to 20 μs are Bogor 1, Bekasi 1, Bekasi 2, Jakarta, and Tangerang 1. The transmitters that get ERP change from 32,62 dbw initial value are Bogor 1, Bogor 3, Bekasi 2 each to 35 dbw, Bekasi 1 to 40 dbw, and Bogor 2 to 28 dbw. Time delay and ERP variations haven t produce the optimal coverage yet. Therefore, the alternative of 2 transmitters as a repeater addition is used. The specifications are, Bogor 1 repeater with 40 m antenna height, 18 dbw ERP, and 10 μs time delay. The other is Bogor 2 repeater with 40 m antenna height, 20 dbw ERP, and 0 μs time delay. With the new specification above, the most optimal SFN coverage for DVB-T service in Jabodetabek area is obtained. VI. ACKNOWLEDGEMENT We would like to thank all relevant personnel in Subdirektorat Penataan Frekuensi Radio Departemen Komunikasi dan Informatika who concede the Chirplus_BC software usage in Ditjen Postel and contribute to the completion of the design. REFERENCES [1], DVB-T Overview, in BC-Training 07a, pp.6, 9-13, 20, LS telcom AG, 2005.